image_id stringlengths 8 28 | caption_L1 stringlengths 51 785 | caption_L2L3 stringlengths 50 1.21k | chart_type1 stringclasses 11
values | cchart_type2 stringclasses 15
values |
|---|---|---|---|---|
ACM_1136.png | This is a scatter plot titled “K-Means Clustering.” The x-axis shows Latitude (11.9 to 12.7), and the y-axis shows Longitude (75.5 to 76.1). The color-coded dots represent three distinct clusters. | The chart shows the purple cluster distribution is concentrated between 76 and 76.6 on the Y axe and spread out along the x axes between 12.3 and 12.7. | Scatter | Simple Scatter |
arXiv_0653.png | This multiple line chart shows the relationship between GPU consumption predictions vs Actual for four different models. The x-axis represents time in seconds, ranging from +0m to +34d 17h, and the y-axis represents GPU usage, ranging from -20.0 to 40.0. | All models closely follow the actual usage trend, but a brief spike around day 17 reveals a divergence where predictions temporarily deviate from actual values before resynchronizing. | Line | Multiple Line |
springer_4259.png | This multiple line chart shows the relationship between the Export Performance and Differentiation advantage. The y-axis shows the Export performance, ranging from 3 to 6. The x-axis shows the Differentiation advantage, ranging from -2 to 2. The chart includes two lines representing "Low competitive intensity" and "Hig... | The chart shows a positive trend, with both High competitive intensity and Low competitive intensity exhibiting a strong upward trend. | Line | Multiple Line |
springer_2561.png | This grouped bar chart shows the "Cell Viability" in percentage on the y-axis in range of 0 to 140 against 5 types of "Concentration" in micro-g per ml on the x-axis: 25, 50, 100, 200, and 400. Three conditions are compared. | The chart shows that the MoO2 condition has the highest median of cell viability, while the HA-MoO2-DOX condition has the lowest. The higher the concentration, the lower the percentage of cell viability. | Bar | Group Bar |
ArXiv_3rd_290.png | This bar chart shows the number of trains on the y-axis in the range of 0 to 8000, against the number of events per train on the x-axis in the range of 0 to 29. | The chart reveals that the number of trains decreases as the number of events per train increases. The highest number of trains is 7500 when there are 3 events per train. | Bar | Simple Bar |
springer_0006.png | This multiple line chart is titled: "JWST Wavefront Monitoring & Maintenance, Cylce 1". The y-axis shows Wavefront Error EMS (nm) on a range from 0 to 100. The x-axis shows the Date from 22-07 to 2023-07. Six conditions are being compared. | The chart shows, a few sharp increases and decreases from 2022-07 to 2022-11, for both the "Piston+Tip+Tilt WFE residuals" and the "Piston+Tip+Tilt AWFE per 48 hours". The "Measured OTE wavefront" conditions stays relatively stable at around 60 wavefront Error RMS with also a few peaks in the beginning. The "Mirror Al... | Line | Multiple Line |
springer_0329.png | This scatter plot is titled "LN vs. LND IPTW", it shows the relationship between the variable on the y-axis, and the SMD on the x-axis in range of 0.0 to 1.0. Two methods are compared as indicated in the legend: Unmatched, and IPTW. | The chart shows a similar trend for both conditions. Most of the data points range between 0 to 0.50 on the x-axis but the Unmatched condition shows two data points over 0.5 with the highest being over 1.0 for "Surgery approach". The PSM condition never gets higher then 0.25. | Scatter | Simple Scatter |
arXiv_0878.png | This bar chart is not titled. The y-axis shows the Degradation (%) in range of 0 to 300%. The x-axis shows three categories: Year 1, Year 2, and Year 3. | The chart shows that the first year has the highest value of Degradation with 235.8%, followed by a significant decrease to 51.4% in the second year, and then a slight increase in the third year to 67.4%. | Bar | Simple Bar |
arXiv_0628.png | This is a line graph titled “Final network proteins response to hypoxia.” The x-axis shows oxygen levels (0.1–0.9 a.u.), and the y-axis shows steady-state protein levels (a.u.). The legend includes red for ILK, cyan for p_473S_Akt, and magenta for p_mTOR. | The graph shows that ILK and p_473S_Akt maintain high steady-state levels under low oxygen but sharply drop around 0.7 a.u., indicating hypoxia sensitivity. p_mTOR shows a smaller decline. | Line | Multiple Line |
springer_4070 copy.png | This bar chart is titled "Joint 1+2 : Mean Absolute Error Comparison (MAE) (Highest Stiffness). The x-axis shows 4 systems, and the y-axis shows the Mean Absolute Error (MAE) in radian ranging from 0.00 to 0.06. | The chart shows that the LQR has the highest mean absolute error, around 0.045, followed by PID. While the SMC has the lowest mean absolute error, around 0.035. | Bar | Simple Bar |
Springer_3rd_104.png | This chart is a heatmap. The x-axis represents the predicted label, and the y-axis represents the True label. Both the x- and the y-axis have the same two categories: '0' and '1'. The cell value represents the count number. | The plot shows that classification has a lower accuracy for the "0" class (791 correctly classified and 507 wrongly classified), but shows a higher accuracy for the "1" class, with 853 correct and 412 misclassified. | Heatmap | Heatmap |
ArXiv_3rd_413.png | This scatter plot with regression line is titled 'Germany', it shows the distribution of shared tweets over time in Germany, with the x-axis representing the year from 2010 to 2020 and the y-axis representing the share of tweets, ranging from 0 to 1.0. The chart includes three different sentiment categories: Negative, ... | The chart reveals that the neutral sentiment dominates in share tweets with a 0.8 value, but shows a gradual decline in the share of tweets in Germany. Both positive sentiment and negative sentiment have lower values in share tweets, between 0 and 0.2. Positive sentiment tweets increase steadily. Negative sentiment re... | Scatter | Scatter plot with regression line |
ACM_0194 copy.png | This horizontal stacked bar chart is titled "Planning to attend graduate school after graduation" and shows the number of responses for two surveys: Survey 1 and Survey 2. The x-axis represents the number of responses, ranging from 0 to 30, while the y-axis lists the two surveys. The bars are divided into two categorie... | In the first survey, 6 respondents indicated that they would attend graduate school after their graduation, while in the second survey, 13 respondents said they would attend graduate school. | Bar | Stacked Bar |
arXiv_0440.png | This horizontal bar chart shows the average time in seconds for two categories: "Single Transfers" and "Batch Transfers". The x-axis ranges from 0 to 3.5 seconds. | The chart shows that single transfers take much more time than batch transfers, having an average time of about 3.6 seconds, compared to batch transfers having 0.2 seconds. | Bar | Simple Bar |
Springer_2nd_1515.png | This is a pie chart titled “Occupation.” It shows the percentage distribution of various occupations: Industry (27.8%), Education or technology (24.9%), Freelance (14.4%), Agriculture (12.6%), Business (10.5%), Legal (5.4%), and Health (4.5%). | The chart reveals that most participants work in Industry and Education/Technology, while Legal and Health sectors are least represented. This suggests a skew toward technical and industrial occupations. | Pie | Pie |
Springer_2nd_1702.png | This heatmap visualizes Average Spatial Attention Weights across various parameters, including head pose, eye ray, and right pose positions and orientations. The x- and y-axes both list the same set of spatial features, forming a square matrix. Color intensity represents weight magnitude, with values ranging from 0.002... | Notably, higher weights cluster near features involving eye_rayDirectionY and rightpose_positionZ, suggesting greater attention relevance in those dimensions. | Heatmap | Heatmap |
arXiv_0103.png | This line chart is titled 'Layer Scales at Last Epoch', it shows the relationship between "Layer Number" on the x-axis ranging from 0 to 30, and the "Layer Scale" on the y-axis, ranging from 0.00 to 0.12. Each data point is marked with a square, connected by a line to indicate the trend. | The chart shows a positive correlation between the layer scales and the layer number. As the layer number increases, the scale also increases. The layer scale shows a sharp increase when layer number is between 20 to 25. | Line | Simple Line |
springer_1794 copy.png | This grouped boxplot displays the expression levels of three genes: DNAJA1, NDUFA5, and SLC2A14, across three groups: Control (red), 16wpi (green), and 24wpi (blue). The y-axis measures expression levels, ranging from 0 to 4. For DNAJA1, expression increases significantly across groups, with asterisks (*** and ****) in... | NDUFA5 and SLC2A14 show smaller but noticeable differences. Outliers are visible as individual dots. The plot suggests gene expression varies over time points (16wpi and 24wpi), especially for DNAJA1. | Boxplot | Grouped Boxplot |
ACM_2071 copy 3.png | This is a line chart titled "Solitary Engagment". The x-axis shows different Monomodal Indicators, such as SP, VB and PY with many sub indicators. The y-axis represents a number, ranging from 0 to 1. | The chart shows that the onlyMonomodal Indicator to have a 1 was "SP.individual.secondary" and"PY.arousal", with all the other ones set to 0. | Line | Simple Line |
ACM_4065.png | This horizontal bar chart compares the percentage of the four metrics in each of the two models. The y-axis lists the four metrics: R squared., MAPE, RMSE(%), and MAE (%). The x-axis represents the percentage values, ranging from 0% to 1%. Two sets of horizontal bars are compared for each medels: XGBost, and Random For... | The chart shows that the XGBoost model outperforms the Random Forest model only in R Squared metric. In the other three metrics, Random Forest slightly outperforms XGBoost model. | Bar | Group Bar |
Springer_2nd_1788.png | This is a scatter plot with bubble size representing values in 2021, plotting the relationship between HDI (x-axis: 0.4 to 0.8) and age-standardized death rates (y-axis: 0 to 90) | Although the chart presents a negative correlation (ρ = -0.36), the p-value of 0.113 indicates statistical insignificance. Countries with higher SDI tend to show lower death rates, but the variability across mid-range SDI values and a few high-outlier bubbles (e.g., Egypt) reduce the strength of this association. | Scatter | Simple Scatter |
springer_2910 copy.png | This scatter shows the relationship between the relative density in percent on the y-axis ranging from 92 to 100 and the VED (J/mms square 3) on the x-axis, ranging from 92 to 100. The chart includes a linear regression line. | The linear regression line shows a positive trend, with a higher relative density at higher VED values. The scatters show less density at lower VED values, with the lowest being around 94%. The highest scatter dot is at 99% density and 1400 VED. | Scatter | Scatter plot with error bars |
0351.png | This bar chart presents the distribution of confidence ratings on a 7-point Likert scale compared of ExampleNet and Online Search. The y-axis shows participants count from 0 to 10, x-axis shows the likert scale. | The median confidence when using ExampleNet is 6, while for participants using online search, the median confidence is 4.5. More pariticiants show confident towards using ExampleNet then Online Search. | Bar | Group Bar |
springer_2503.png | This bar chart shows the frequency of p-values. The x-axis represents the p-values, ranging from 0 to 0.14, while the y-axis represents the frequency, ranging from 0 to 1. There is a line titled "Average p-value" with the value 0. | The chart reveals that there is one peak at 0 p-value with a frequency of nearly 1 and the rest of the graph is empty. | Bar | Simple Bar |
arXiv_0717.png | This grouped boxplot presents the Normalized Preference Score across six dimensions: Indiv/Collect., Indulg./Restraint, Long/Short Term Orient., Masc/Fem., Power Dist., and Uncert. Avoid. Index. The comparison is between two rating types: Comparative Rating (blue) and Absolute Rating (orange)—with each boxplot showing ... | The plot reveals that the "Competitive" dimension has higher median values and narrower interquartile ranges, suggesting higher perceived negative feedback scores. In contrast, the "Absolute" dimension shows lower median values and wider interquartile ranges, indicating more varied feedback ratings. | Boxplot | Grouped Boxplot |
springer_0146.png | This scatter plot with regression line is not titled. It shows the Pr(%) on the y-axis ranging from -2 to 0. The x-axis shows the Phase angle a (°) on a range from 0 to 7. There are three conditions compared: TNOs, Chiron and Pholus. | The plot shows four lines. One stable at 0, one ranging from 0 to -1.5 with a steep decrease where the TNOs data points range around. The Chiron points range around the line from -1.5 to -1 and the Pholus has the lowest Pr with points ranging around -2. | Scatter | Scatter plot with regression line |
ArXiv_3rd_232.png | This grouped bar chart shows the mean Lacyency on the y-axis in the range of 0 to 1000, against the number of workflow instances on the x-axis in the range of 10 to 500. Three conditions are compared as indicated in the legend: VCEFlex, VELA, and VECA. | The chart reveals that the VCEFlex condition consistently outperforms both VELA and VECA across all workload instances. The VELA condition shows a decreasing trend with the increase of the number of workflow instances, while the VECFlex and VECA conditions show constant mean latency across the number of workflow instan... | Bar | Group Bar |
extra2_0318.png | The chart is a pie chart titled "Unimpacted". It includes a legend with five categories: dispersal filter, competitive exclusion, environmental filter, unexplained absences and NA. The segments represent proportions of outcomes. | The chart reveals that 94.9% of outcomes are due to dispersal filters, highlighting a dominant trend. The remaining categories indicate varied but lesser proportions, such as "competitive exclusion" with 1.9%. | Pie | Pie |
ACM_3170 copy.png | This scatter plot with regression line is titled "Abstract Score vs. Comprehensive Score". The x-axis shows the "Writing Score" ranging from 3.5 to 7.0, and the y-axis shows the "Comprehensive Score" ranging from 0 to 10. | The plot shows a positive trend between abstract and comprehensive score, with a linear regression line indicating a strong positive correlation. The regression line slopes up steeply, showing a significant increase in abstract scores as comprehension scores rise. | Scatter | Scatter plot with regression line |
springer_0629.png | This multiple line chart shows the relationship between reaction time (min) on the x-axis and the HMF yield (percentage) on the y-axis, ranging from 0 to 35%. The chart includes three lines representing different temperatures: 413 K, 423 K, and 433 K. | The chart shows a peak in HMF yield for 413 and 423 K at 250 minutes. 423 has the highest peak of all measured temperatures, 30% yield. 433K has a peak at 200 minutes, this is lower than the peak of 423 K but higher than the peak of 413 K. | Line | Multiple Line |
springer_1601 copy.png | This scatter plot displays mean values for candidate favorability and vote choice under the Dam scenario across four reporting treatments: Control, Investigative, Event-Oriented, and Combined Reporting. The x-axis lists outcome categories, while the y-axis ranges from 0.25 to 0.75. | Combined Reporting (black squares) consistently results in higher mean values, especially for Support for Bond and Challenger Favorability. Control and Investigative treatments yield the lowest means, particularly for Incumbent Favorability, highlighting a general shift toward challenger support in response to Combined... | Scatter | Simple Scatter |
Springer_2nd_1351.png | This heatmap is titled 'Q-Values Across Parameter Variations'. The y-axis lists six parameter variations: Low or High for α, γ, and ε. The x-axis lists four states: CC, CD, DC, and DD. The cell value represents Q-values, ranging from 20 to 100, as indicated on the right side. | The chart shows that State CC consistently has the highest Q-values across all parameter settings, peaking at 100 with High γ. DD state shows the lowest values throughout. Q-values improve with increased α and γ. | Heatmap | Heatmap |
1494a.png | This bar chart is titled 'All Sample', it compares the "Perceived Brightness" for two categories: "Black" and "White". The y-axis represents perceived brightness, ranging from 0 to 240. Each box represents the distribution of perceived brightness, with the median line, interquartile range (IQR), and whiskers showing th... | The plot shows that the "Black" samples have a higher perceived brightness than the "White" samples. The median perceived brightness for "Black" is around 170, with most data points falling between 140 and 200. In contrast, "White" samples have a median around 130, with a wider spread from 60 to 180, indicating greater... | Boxplot | Simple Boxplot |
arXiv_1035.png | This is a heatmap sows three models on the y-axis: DeepSeek-R1, OpenAi o3-mini, and Salamandra. Eight types of responses are shown on the x-axis. Each cell is displaying the percentage of response to the model. A color scale ranging from 0 to 100 is displayed on the right side. | DeepSeek R1 shows the minimal responses across all categories. OpenAI o3-mini displays moderate level of responses, and Salamandra exhibits consistently higher responses percentage, especially in Profession (63.6%) and Race & Nationality (56%) | Heatmap | Heatmap |
extra2_0271.png | The chart is a pie chart titled "Overall Parents' Preferences." It features a legend with five categories. The pie sections display percentage values. | The chart reveals that a significant majority of parents prefer willingness to spend extra expense on program at 32.36%. In contrast average daily exercise time shows minimal preference at 6.85%. | Pie | Pie |
pie_48.png | The chart is a pie chart titled "Two-thirds of Hispanic adults said being Hispanic was part of their racial background." It includes a legend with categories: Racial background, Ethnic background, Both, and Don't know. Each category is represetned by one slice of the pie. | Most Hispanic adults (67%) view being Hispanic as part of their racial background, with 56% considering it both racial and ethnic. Smaller segments identify it solely as ethnic (19%) or racial (11%). | Pie | Pie |
arXiv_0473.png | This multiple line chart is titled with "Cybersecurity Publications in Banks (Post-Industry 4.0 Era)". The x-axis shows three 5-Year periods, and the y-axis shows the Number of Publications ranging from 0 to 2000. 4 categories are compared as indicated in the legend. | The chart shows that the number of publications increases across three categories, particularly for "Springer", which has more than 500 "Number of Publications" in the 2016-2020 period and over 2000 in the 2021-2024 period. Only category "ACM" has across all three periods a "Number of Publications" of 0. | Line | Multiple Line |
springer_3097 copy.png | This stacked bar chart illustrates the relationship between the quantities of cobalt from 2024 to 2040. The x-axis represents the years from 2024 to 2040. The left y-axis indicates the quantities of cobalt (10 thousand units) ranging from 0 to 50. the right y-axis represents the percentage of the rate ranging from 0 to... | The chart shows a steady increase in cobalt demand over time, accompanied by growing contributions from both RR and GR, with the overall recovery rate projected to exceed 55% by 2040. | Bar | Stacked Bar |
ACM_2273.png | This multiple line chart shows Accuracy curve for different epochs. The x-axis represents epochs in the range of 0 to 120, the y-axis shows accuracy percentage ranging from 50 to 90%. 5 other models are compared as indicated in the legend. | The chart shows that all five models perform similarly in accuracy, with the DPP match leading at 90% accuracy. The models' accuracy improves with epochs, with the WideResnet showing the least percentage of accuracy at the end of the epoch trend. | Line | Multiple Line |
arXiv_0085.png | This multiple line chart shows the relationship between the performace on the y-axis, ranging from 0 to 0.6, against the percentage on the training data on the x-axis, ranging from 5% to 100%. The chart includes two lines representing "ISTM" and "LSTM with DDA," as indicated in the legend. | Both conditions show a positive trend between performance and training data percentage. The conditio 'LSTM with DDA' shows a slightly higher increase compared to the 'Istm' condition. | Line | Multiple Line |
springer_4204 copy.png | This boxplot is titled 'Female Participants Overall Accuracy'. It shows the value of accuracy on the y-axis from 91 to 96. The x-axis shows one boxplot representing female participants. | The chart shows that the median accuracy for female participants is 94%, with most values concentrated between 92.5% and 95.5%. The lower whisker extends to around 91.5%. | Boxplot | Simple Boxplot |
pie_143.png | The chart is a pie chart titled "Russian Support of Assad Regime." It includes a legend with three categories: Support, Oppose, and DK (Don't Know). Each category is represetend by one slice of the pie. | The largest segment, 52%, represents those who don't know. Support is 27%, slightly higher than the 22% who oppose. | Pie | Pie |
springer_1375 copy.png | This scatter plot with error bars shows the proportion of synthetic future usage across three sentential polarity types: affirmative, negative with passive, and other negative word. The x-axis represents categories (n = 3,291), and the y-axis ranges from 0.00 to 1.00. | Usage is low for affirmative and passive-negative sentences (~0.20), but significantly higher (~0.65) for other negative contexts. | Scatter | Scatter plot with error bars |
extra2_0294.png | The chart is a multiple line graph titled "Income effect," displaying two variables: Income and Income x combined value. The x-axis represents different stages (First contact, Top 1 Fit, Expert diagnosis, Interview), while the y-axis indicates income in euros, ranging from 1,000 to 2,000. | The graph reveals that income is consistently higher than income x combined value across all x-values, and both are highest for "Interview." | Line | Multiple Line |
springer_1948.png | This multiple line chart shows the Extinction on the y-axis, against the Wave length in nanometers on the x-axis in range of 400 to 900. Three lines represent three conditions: AuNP, AuNP/CS, and AuNP/CS-Res. | The line chart shows a general decline in extinction as wave length increases for all three models. The AuNP model is slightly higher than the others until all three sharply peak at wavelength 525, then AuNP/CS is slightly higher than the other two. | Line | Multiple Line |
springer_3080 copy.png | This multiple line chart shows the "decomposition voltage (V)" on the y-axis, ranging from 2.0 to 5.5, and "Temperature (C)" on the x-axis, ranging from 800 to 1200. The chart includes five lines representing different conditions: Ca2F, LiF, NdF3, NaF, KF, KCI, LiCl, CaCl2, NaCl, KCI, Nd2O3. | The chart shows a general decline in value across all temperatures, with the highest values at lower temperatures. The CaF2 technique consistently outperforms all the other methods across all temperatures, with the highest values at lower temperatures. The lowest values acheived by Nd2O3. | Line | Multiple Line |
Springer_2nd_0309.png | The chart is a horizontal bar graph. It includes four generations: 1st, 2nd, 3rd, and 4th. The x-axis shows percentage scales from 0% to 80%, while the y-axis lists the generation names. | The chart shows the 2nd generation at 68% being the most significant, followed by the 3rd generation at 24%. The stark contrast with the 1st generation at 8% highlights varying representation among the generations. | Bar | Simple Bar |
extra2_0299.png | The chart is a line graph titled "Group of patients with positive income effect." It features two lines, with a legend indicating "Income" and "Income x combined value." The y-axis is labeled "Income" and ranges from 700 to 2,200 euros, while the x-axis denotes four categories: First contact, Top 1 Fit, Expert diagnosi... | The graph illustrates trends in income across different evaluations, showing a significant increase in both income and combined value at the interview phase. both values start around 1200 euros and end around 2200. | Line | Multiple Line |
extra2_0275.png | The chart is a pie chart titled "Preferences Differences in Father and Mother Respondents." The legend identifies attribute categories, and the pie segments represent percentages for each category. | The chart reveals that a significant majority of mothers prefer willingness to spend extra expense on program at 32.33%. In contrast Average Daily Exercise Time shows minimal preference at 2.81%. | Pie | Pie |
1770.png | This clustered bar chart displays the number of preregistrations ranging from 0 to 45 on the y-axis, compared across eight categories of preregistered information along the x-axis. The categories include A-priori description, Research questions/hypotheses/outcomes, Variables, Design plan, Analysis plan, Sampling plan, ... | Statistics: The highest number of available preregistrations appears in the Design plan (44), while the lowest number is in the Analysis script category (14). Comprehensible preregistrations are highest in the A-priori description and Analysis plan categories (36 and 33, respectively). Consistent preregistrations are h... | Bar | Group Bar |
arXiv_0555.png | This is a horizontal bar graph titled “Number of data instances in each held-out disease dataset.” The x-axis shows “Number of Cells,” ranging up to 300,000, and the y-axis lists five diseases. | Rheumatoid Arthritis has the highest number of cells, followed by Chronic Kidney Disease and Gastric Cancer. Alzheimer’s Disease and Pediatric Crohn’s Disease have the lowest cell counts. | Bar | Simple Bar |
arXiv_0336.png | This multiple line chart is titled DS1 and it shows the estimated MLL (nats) on the y-axis, ranging from -7300 to -7150. The x-axis represents the iteration from 10 square1 to 10 square 5. The chart includes five lines representing different models: VBPI10, VBPI20, LOOR, REP and VIMCO. | The estimated MLL values for all models initially increase with the number of iterations. The REP model is the most stable, reaching the highest value with minimal fluctuations. In contrast, the VBPI20 model shows the most pronounced fluctuations, including the largest spike shortly after 10² iterations. All conditions... | Line | Multiple Line |
ArXiv_3rd_35.png | This bar chart is titled "Country of origin of expert survey participants" and shows the number of respondents from various countries around the world. The y-axis represents the number of respondents, ranging from 0 to 20, and the x-axis lists the countries, including France, Mauritius, Nigeria, Cameroon, Argentina, Et... | The chart shows that Nigeria has the highest number of experts (16), followed by Kenya (13). Other countries have a very low number of survey participants, below 5. | Bar | Simple Bar |
extra2_0109.png | The chart is a box plot titled "Text Entry: Error Rate." The y-axis represents Character Error Rate (CER) with a scale from 0.00 to 0.20, while the x-axis categorizes the groups as Baseline, Overlay Normal, and Overlay Large. | The chart reveals that the Baseline condition consistently exhibits the lowest error rates compared to Overlay Normal and Overlay Large (0.03 vs. around 0.1). The significant difference suggests a notable impact of overlay size on performance, especially error rates. | Boxplot | Simple Boxplot |
ACM_4291 copy 2.png | This multiple line chart is titled "T right arrow I@Colorectal 16 bit". It shows text-to-image retrieval performance on the Colorectal dataset using 16-bit hashing. The x-axis represents Recall (0.0–1.0) and y-axis shows Precision (0.2–0.8), 10 methods are compared. with each curve labeled in the right legend using uni... | All methods show decreasing precision with increasing recall, but “Ours” method consistently outperforms others across the entire range. | Line | Multiple Line |
3143.png | This boxplot is titled 'Mean Absolute Error of Two Conditions'. It shows Mean Absolute Error(N) on the y-axis ranging from 0 to 8, the x-axis lists two conditions: Staitc condition and synamic condition. | Dynamic condition shows higher median dor mean absolute error thean static condition. However, static condition has wider range of distribution than dynamic condition. | Boxplot | Simple Boxplot |
ACM_1223.png | This is a multiple line graph titled "% VETT beats local (FH=192, Internal)". The x-axis represents Target time-series available, ranging from 0 to 0.8, and the y-axis represents the percentage of VETT better or Local better, ranging from 0% to 100%. The legend exhibits 4 datasets in different color dots. | The graph demonstrates that VETT generally outperforms local models across datasets, especially when less target data is available. Performance varies by dataset, with Energy consistently showing the highest advantage.The horizontal dashed line at 50% indicates the threshold where the models outperform the local approa... | Line | Multiple Line |
Springer_2nd_1605.png | This is a line graph titled “Manganeses.” The x-axis shows six labeled points (Point 1 to Point 6), and the y-axis shows values from -0.04 to 0.04. A red line with circular markers represents the measured values, all of which are 0. | The graph demonstrates that manganese levels remained constant at zero across all six data points. This flat trend suggests no detectable presence or variation in manganese concentrations during the measured intervals. | Line | Simple Line |
ACM_1854.png | This bar chart is titled "Distribution of Bootstrapped AUCDifferences (Female - Male)". The x-axis shows the AUCDifference (Female AUC - Male AUC) on a linear scale from 0 to 0.1. The y-axis shows Frequency on a linear scale from 0 to 600. | The AUCDifference peaks around 0.03 at a frequency around 550. | Bar | Simple Bar |
1142.png | This grouped bar chart, titled "Assessment of GUI Elements for Accessibility," displays the percentage of each GUI element categorized under Supporting (S), Partially Supporting (PS), and Not Supporting (NS) accessibility standards across various tools. The y-axis is labeled Percentage, with values ranging from 0% to 1... | Notably, TextField, ToggleButton, Menu, and Menu Button exhibit the highest "S" percentages, indicating strong support for accessibility. Conversely, Group and Dropdown display the highest "NS" percentages, suggesting they are least supportive of accessibility measures. | Bar | Group Bar |
springer_4469.png | This line chart shows the relationship betweenTransmittance and Wacelength. The y-axis shows Transmittance in percent and the x-axis Wavelength in centimeters to the power of -1, ranging from 4000 to 500. The chart includes three lines, named a, b and c. | The chart shows an overall decline in transmittance as the wavelength gets smaller. The line c has the highest transmittance, followed by b and then c. | Line | Multiple Line |
springer_4052.png | This grouped bar chart shows the impact scores for three policies under two conditions. The x-axis represents three different policies: Policy A, Policy B and Policy C. The y-axis represents the impact score, ranging from 0 to 70. The chart compares two conditions as indicated in the legend: Investment Impact and Techn... | The chart shows that Policy A has the highest impact score over both conditions, followed by Policy B and C. The condition Technological Adoption is the highest for Policy and B, and the Investment Impact for Policy C. Policy A has the greatest difference between the two conditions. | Bar | Group Bar |
springer_2554 copy.png | This multiple line chart is titled "C" and shows the percentage of Yt on the y-axis, ranging from 0% to 70%, against Time on the x-axis, ranging from 0 to 240 minutes. The chart includes three lines representing different conditions: 10%, 5%, and 2.5%. | The chart shows that all conditions increase in percentage as time increases. The 10% condition has a higher percentage than the other conditions. | Line | Multiple Line |
ArXiv_3rd_288.png | This heatmap shows the joint distribution of “MC” (x-axis) and “BP” (y-axis), both scaled from -1.0 to 1.0. A vertical color bar labeled “count”, ranging from 100 to 260, is displayed on the right side. | The chart shows the counts concentrate around (MC is 0.00, BP is -0.25), forming a dense central region. | Heatmap | Heatmap |
springer_4026.png | This grouped bar chart is titled 'Distribution of Cosine Similarity Scores for Path Search Methodologies'. The y-axis shows Frequency in range of 0 to 25, the x-axis shows Cosine Similarity in range of 0.00 to 0.30. As indicated in the legend two Path Search Methodologies ar compared: "All Shortest Paths" and "Template... | The chart shows that the "Template-Based" category has a notably higher median of frequency compared to the "All Shortest Paths" category. The "All Shortest Paths" category has a bigger range in Cosine Similarity than the "Template-Based" category, from under 0.00 to over 0.30. | Bar | Group Bar |
springer_3629 copy.png | This boxplot shows the value of observed features for three sites. The y-axis represents the value of observed features, ranging from 4200 to 5000. The x-axis lists the three sites: Site 1, Site 2, and Site 3. | The plot shows that Site 2 has the highest median value of observed features, around 4650. Followed by Site 3. Site 1 has the lowest median and narrowest inter-quantile range in the value of observed features. | Boxplot | Simple Boxplot |
arXiv_0359.png | This bar chart shows the F1 scores for various conditions. The y-axis lists the conditions, including "RN-Net," "Iforecast," "Autoencoders," "DAGMM," "elliptic envelope," "DevNet," "GAN," "MGBTAI," "PReNet," "Deep-SAD," and "FTTransformer." The x-axis represents the numbers of the F-1 scores, ranging from 0 to 50. | The chart shows that "RN-Net" has the highest F1 scores, followed by "FTTransformer" and "Deep-SAD". "LOF", "Elliptic Envelope" and "DevNet" have the lowest and very similar F1 scores. | Bar | Simple Bar |
springer_4571.png | This line chart shows the relationship between time and power. The x-axis represents the tim in hours, minuts and seconds, ranging from 8.38.23PM to 8.58.46PM. The y-axis represents the power in watt, ranging from 0 to 100. Four lines are compared as indicated in the legend. | The chart shows that the LoadB condition is the moast inconsistent. The HigherPoint and LowerPoint condition stay the same throughtout the time. | Line | Multiple Line |
extra_0177.png | This heatmap is titled "Proximity to Waterbodies". It shows the results through a map. Latitude (24°59′N to 25°16′N) is shown on the y-axis, and longitude (75°41′E to 75°59′E) is listed on the x-axis. The color bar is displayed on the right side, indicating elevation ranging from Low: 0 to High: 0.069. | The chart shows that the low proximity to waterbodies is dominant in this entire region. The medium and high proximity to Roads only appears sparsely in the south and west regions. | Heatmap | Heatmap |
arXiv_0077.png | This multiple-line chart shows the value of p in the function of s on the y-axis, ranging from 10 power of -6 to 10 power of 1. The x-axis shows the value of s ranging from 0 to 10 power of 3. Three conditions are compared: 't=95min', 't=189min', and 't=340min', as indicated in the legend. | All conditions follow a downward trend. When s value is larger than 10 power of 1, there is large deviation in the value of p for three conditions. | Line | Multiple Line |
ArXiv_3rd_664.png | This multiple line chart shows "number of ENVs" on the left y-axis from 0 to 7, and "ENV_0 fraction" on the right y-axis from 0.0 to 1.0. The x-axis shows the changes in Time resolution Delta t. | The number of ENVs remains constant at 5 for fine time resolutions, then drops sharply to 1 and finally to 0 as resolution increases. Meanwhile, the ENV₀ fraction increases and rapidly reaches 1.0 | Line | Multiple Line |
arXiv_1030.png | This is a heatmap with a title 'N = 15'. The x-axis is "Exploration Rate (T)" from 0 to over 6, and the y-axis is "Edge Probability (p)" from 0.1 to 1. A colorbar labeled "Value" ranging from 0 (dark blue) to 1 (yellow) is shown on the righ. | High values dominate when edge probability is high and exploration rate is low, forming a sharp transition boundary around T ≈ 1–2. | Heatmap | Heatmap |
ACM_2665 copy 3.png | This line chart is titled Mixed Signal 2, shows Amplitude on the y-axis in range of -200 to 200 against Sampling points on the x-axis in range of 500 to 2500. | The chart shows that the amplitude of the signal fluctuates over time, with intensity at 0 Aplitude. | Line | Simple Line |
0112b.png | This bar chart, titled "False-discovery Rate (FDR) of Two Groups," compares the FDR for "Prior Elicitation" and "Control" groups. The x-axis represents the two groups, and the y-axis shows the FDR, ranging from 0 to 0.5. | The chart shows that the "Control" group has a slightly higher FDR compared to the "Prior Elicitation" group, indicating a marginally higher rate of false discoveries without prior elicitation. The difference between the groups suggests that prior elicitation may contribute to a modest reduction in FDR, potentially imp... | Bar | Simple Bar |
arXiv_0426.png | This heatmap shows the upper X squared maximum value for each combination of Theta on the x-axis ranging from 1 to 100, and alpha on the y-axis ranging from 13 to 7. The legends shows that upper X squared maximum ranges from 2000 to 8000. | The chart shows that almost all values are under 1000, except when alpha is 7 and Theta is 40 or higher, There is a clear increase in upper X squared maximum when Theta increases, and also when alpha decreases. The highest value is for Theta 100 and alpha 7, where the upper X squared maximum is 9810.6. | Heatmap | Heatmap |
extra_0151.png | This boxplot shows CH4 concentration (ppm) on the y-axis from 1.8 ppm to 2.6 ppm. The x-axis lists 10 categories: six sites (MA, SG, GZ, SM, LM) across the summer and autumn seasons. Each category is represented by one boxplot. | The chart shows that the LM autumn category has the highest CH4 concentration, around 2.5 ppm, and the LM summer category has the lowest median in CH4 concentration, around 1.9 ppm. | Boxplot | Simple Boxplot |
springer_2510.png | This bar chart is titled "Obese vs. Severely Obese" and shows the frequency of the number of ATEs. The x-axis represents the ATE values, ranging from -0.02 to 0, while the y-axis represents the frequency, ranging from 0 to 0.12. There is a line titled "Average ATE" with the value 0.006. | The chart reveals that the distribution of frequencies is concentrated between -0.01 and 0. The average frequency is also significantly higher than the others, as indicated by the red dashed line at 0.006 ATE and a peak frequency of 0.12. | Bar | Simple Bar |
0205a.png | This is a box plot illustrating "Sadness Changes" for two groups: "Wearer" (n=7) and "Non-wearer" (n=14), comparing two conditions: "Glasses without tears" and "Glasses with tears". The y-axis is labeled "Sadness Changes" and ranges from -8 to 8. Significant differences are indicated by an asterisk (*p < 0.05). | The chart shows a clear pattern of increased sadness in both "Wearer" and "Non-wearer" groups when glasses are with tears, with significant differences marked between the "Glasses without tears" and "Glasses with tears" conditions. The box plot for "Glasses with tears" consistently displays higher median values in both... | Boxplot | Grouped Boxplot |
ACM_1855 copy 2.png | This scatter plot with error bars is titled 'Total Dataset Size = 10000 Class Distribution = 0.3, Error Rate = 0.1'. The x-axis shows Target Group Size in range of 10 to 5000, the y-axis shows FPRDifference in a range of -0.6 to 0.6. | The plot shows that the FPRDifference fluctuates between -0.1 and 0.1 until the Target Group Size increases from 10 to 500, then stays at zero. The size of the error bars goes down from around a 0.3 total spread to around 0 after 2000 target group size. | Scatter | Scatter plot with error bars |
1699.png | This line chart represents the force-displacement relationship of hemisphere samples at three different internal air pressures: 100 KPa (blue line), 200 KPa (green line), and 300 KPa (purple line). The x-axis shows the displacement in millimeters (mm), ranging from 0 to 12 mm, and the y-axis shows the force in Newtons ... | The chart shows that as the displacement increases, the force also increases for all three pressure levels. The force required at 300 KPa increases more rapidly compared to 100 KPa and 200 KPa, indicating that higher internal air pressure results in higher force at a given displacement. | Line | Multiple Line |
springer_3894.png | This line chart shows the relationship between the number of initial phi points (x-axis) and the deltaPH (y-axis). The initial pH is ranging from 0 to 12, and the Dealta PH values range from -1.0 to 2.5. There is an arrow points to the intersection with y-axis eqauls 0, labelled PZC=6.68. | The Delta pH drops to -0.8 when inital pH is 5, and then it increases to 2.0 when initial pH is 10, and then dealta pH drops to 0.5 when initial pH is 12. | Line | Simple Line |
2088b.png | This line chart compares the "Distance Response" of four devices—Accelerometer, Geophone, Traditional Microphone, and Voice PickUp (VPU)—with "Distance (cm)" on the x-axis, ranging from 5cm to 130cm and "Relative Response (dB)" on the y-axis, ranging from 0 to -12 dB. Each device's distance response is represented by a... | All devices show a decrease in relative response as distance increases, though the rate of decline varies. The geophone and accelerometer show the most significant drop, especially beyond 15cm. The traditional microphone and VPU show more gradual decreases. | Line | Multiple Line |
springer_4053.png | This grouped bar chart shows the demand for hydrogen across six years. The x-axis represents the years from 2024 to 2049. The y-axis represents the hydrogen demand in million tons, ranging from 0 to 10. The chart includes six colored bars, each representing a different country: Egypt, Saudi Arabia, Libya, Jordan, Franc... | The chart shows that the demand for hydrogen increased steadily from 2024 to 2049. The highest demand for hydrogen is in 2049, the lowest in 2024. The country with the highest demand is France over all six year, followed by Italy and then Saudi Arabia. The lowest demand over all six years is in Jordan. | Bar | Group Bar |
ArXiv_3rd_394.png | This multiple line chart shows Count on the y-axis ranging from 10 to the power of 0 to the power of 5, against Distance on the x-axis ranging from 0 μm to 1500 μm. Two conditions are compared, as indicated in the legend: Brine and Gas. | The chart shows that Brine exhibits a much higher initial count and steeper decay with distance, while gas starts lower but has a slower decline. There are no more counts for Brine after 800 distance, and there are no more counts for Gas after 1000 distance. | Line | Multiple Line |
springer_3612 copy 2.png | This scatter plot shows the percentage of Pupae weight on the y-axis from 0 to 1.4, against the Soil depth (1-6cm) on the x-axis from 1 to 6. The annotation shows the fitting line is y=--0.0512x2 + 0.1583x + 1.1109, with R square =0.964. | The percentage of Pupae weight remains high at lower soil depth values and decreases steadily as soil depth increases. The strong R² indicates a very good quadratic fit | Scatter | Scatter plot with regression line |
0540.png | This scatter plot displays Median PABAK in range of 0 to 1.0 on the y-axis agains Focus Window (m) on the x-axis in range of 0 to 60m. 4 types of Consine Distance Threshold are compared: 0.1, 0.2, 0.3, and 0.4. | Lower distance thresholds demonstrate consistently higher PABAK values compared to higher distance thresholds. Additionally, longer focus time windows contribute to increased PABAK values. | Scatter | Simple Scatter |
ACM_2232.png | This bar chart displays the average Intensity on the y-axis in range of 0.00 to 0.15 against 10 topics on the x-axis. Two conditions are compared for each topic: Intensity threshold and Topic intensity. | The chart shows that topic intensity is higher than Intensity threshold in Topic Intensity one, three, four, six and eight conditions. | Bar | Simple Bar |
0246.png | This Boxplot shows the cognitive load of participants using DeepLens and the Baseline tool. The cognitive load is measured on x-axis in five different dimensions, including mental demand, Hurry, Performance, Effort and Frustration. 1-7 Likert Scale is used on the y-axis. | For "Mental Demand" and "Frustration," participants using DeepLens report lower medians and narrower interquartile ranges . "Performance" is rated higher by participants using DeepLens. "Hurry" and "Effort" appear similar across groups | Boxplot | Grouped Boxplot |
extra_0656.png | The chart is a heatmap. The legend indicates values are encoded with a gradient to show low and high values. The x-axis and y-axis are labeled with categorical variables including 'sex' and 'age', with cells displaying values ranging from 0.2 to 0.7. | The chart reveals a strong correlation among certain variables, particularly the highest value of 0.7 between age and P1. Comparatively, the weakest relationship is noted with a value of 0.2, P2 and D1. | heatmap | heatmap |
1017.png | This horizontal bar chart is tittled "Comparison of ASD Participants' Technological Preference for Managing Stress". Technologiral Preferences are list on the y-axis, and the percentage of respondents are shwon on the x-axis in range of 0% to 80% | The top choices include using a smartphone, headphones for noise cancellation, headphones for music, and a weighted blanket. | Bar | Simple Bar |
ACM_3034.png | The x-axis on this line chart represents time in seconds (0 to 0.5 seconds), and the y-axis represents the amplitude from -1 to 1. | The chart reveals a general strong fluctuation over time and a light downward trend. | Line | Simple Line |
Springer_2nd_1537.png | This is a histogram titled “D: Histogram of Sand Content (Kg/m³).” The x-axis shows sand content in Kg/m³, and the y-axis shows count. A pink density curve overlays the bars; a legend displays 9 statistical details. | The histogram shows sand content distribution across 36 samples, with most values clustered around 614.83 Kg/m³. The slight negative skew and low kurtosis suggest a mildly left-tailed and flat distribution. | Bar | Group Bar |
springer_0938 copy 3.png | This scatter plot is titled 'males' and shows test score PRE on the x-axis, ranging from 0 to 10, against confidence score PRE on the y-axis, ranging from 1 to 4. | The plot reveals a positive relationship between test scores PRE and confidence score PRE, indicating that test score increases as confidence score increases. | Scatter | Simple Scatter |
springer_4214.png | This boxplot shows the mean square error for four methods. The y-axis represents the four methods: MARS, MLP, SVM and DT. The x-axis shows the mean square error in range of 10 to the power of -5 to 10 to the power of -1. | The chart shows that the SVM has the highest mean square error, meaning it makes large prediction errors. The DT has the lowest mean square error, followed by MARS, which means it's the most accurate and consistent. | Boxplot | Simple Boxplot |
extra_0014.png | This is a heatmap chart titled "Pinch." The legend on the right indicates the color scale for values ranging from -50 to 0. The x-axis is labeled "Time" and spans from 0 to 140, while the y-axis is labeled "Doppler," ranging from -400 to 400. | The chart illustrates the Doppler values fluctuating around 0 with a rhythmic motion and high-intensity values across time. The intensity values are low at other Doppler values. | Heatmap | Heatmap |
springer_4266.png | This line chart shows the relationship between "Value-based Pricing" on the x-axis and "Profitability" on the y-axis, with two lines representing "Low CPC" and "High CPC". The chart includes a legend indicating the line type. | The plot shows that low-CPS pricing leads to higher profitability compared to high-CPS pricing. | Line | Multiple Line |
springer_0673.png | This horizontal bar chart shows the importance of various features. The y-axis shows different economic features, the x-axis shows the importance in percentage, ranging from 0 to 10% | The most important feature is investment with 9.73%, followed by inflation with 9.73%. The least important feature is 'wealth gini', with only 3.93%. | Bar | Simple Bar |
ACM_0100.png | This multiple line chart shows Active Groups Ratio Trend by method. The x-axis shows session ID from 1 to 11, the y-axis shows Active Groups Ratio ranging from 0.0 to 1.0. Two methods are compared as indicated in the legend: random, and expertise balanced. | The chart shows a decreasing and fluctuating trend in both categories. Both methods had a drastic fall in session ID 8. In session 11 both methods are around 0.4 active groups ratio. | Line | Multiple Line |
arXiv_0175.png | This heatmap is titled 'Consistency map (Downsampled) Ops_2'. It shows time in seconds on the x axis, ranging from 7000 to 8500 seconds. Repeat number is shown on the y-axis, ranging from 0 to 50. The heatmap shows voltage for each combination of time and repeat. | The map shows a voltage close to 0 for all values, except at time instances just after 7500 sec, 7750 sec, 8000 sec and 8250 sec. Then, the voltage decreases briefly to -0.03 V, and increases afterwards to 0.01 Volts, and returns to the basis level of approximately 0 Volt after. This happens for almost all repeat numbe... | Heatmap | Heatmap |
ArXiv_3rd_542.png | This multiple-line chart is titled "DADBS performance under different attack scenarios”. Values of System Performance are plotted on the y-axis from 50% to 100%, against Percentage of Malicious Nodes on the x-axis from 0 to 50. Two conditions are compared and listed in the legend: Throughput and Consistency. Each cond... | The chart shows that both conditions have a similar and decreasing trend, but the "Throughput" condition decreases more rapidly as the percentage of malicious nodes increases than the 'Consistency' condition. | Line | Multiple Line |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.