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 |
|---|---|---|---|---|
springer_2088.png | This grouped bar chart shows the percentage of growth inhibition on the y-axis ranging from 0 to 100, and four groups on the x-axis. Two conditions for each group are indicated in the legend: BPS-ZnO-NPs, and Zn(OH) 2. | The bar chart shows that the BPS-ZnO-NPs condition resulted in significantly higher growth inhibition in all groups. The S. aureus, E. coli and C. albicans groups have the highest growth inhibitions, and A. niger has the lowest. | Bar | Group Bar |
springer_3140.png | This bar chart is titled the "Friendship synchrony Publications". On the y-axis, it shows the number of publications ranging from 0 to 7, and the "Year of Publication" on the x-axis, ranging from 1982 to 2022. Each bar represents the number of publications for a specific year. | The number of publications is at 1 in the first two years. Then it rises from 95 to 98 to two publications. After that from 2001 to 2018 the publications are again at 1. then it shows a slight increasing trend, with the most publications in 2022 (6). | Bar | Simple Bar |
pie_88.png | The chart is a pie chart titled "52% of STEM-trained college graduates are employed in the STEM workforce." It includes a legend with categories: STEM, Management/business/finance, Other non-STEM, and Social services/legal/education. Each category is represetend by one slice of the pie. | STEM employment is the largest segment at 52%, followed by Other non-STEM at 20%. Management/business/finance accounts for 17%, while Social services/legal/education is the smallest at 11%. | Pie | Pie |
ACM_1435.png | This scatter plot with regression line is titled “Access control by RUT (Qp not equal to Qq).” The X-axis shows predictions from the sparse Gaussian Process model, ranging approximately from -1 to 2.0. The Y-axis represents predictions from the secure sparse Gaussian Process model, with a similar range. A dashed diagon... | Data points appear scattered without a clear trend along the diagonal, suggesting noticeable differences between the two models’ predictions. | Scatter | Scatter plot with regression line |
extra_0548.png | This pie chart is titled 'e.Carbon neutral scenario. It shows the breakdown of three categories: Low, Medium, and High. Each category is represented by one segment of the pie, with the percentage labelled. | The chart shows that the 'Medium' category has the largest percentage, with 50.11%. The 'Low' category follows, with 41.18%. The 'High' category has the least percentage, of 8.70%. | Pie | Pie |
Springer_2nd_1435 copy 2.png | This stacked bar chart shows the Quantity of cadmium discharge (Kg) on the y-axis, from 0 to 2500 Kg. The Year is on the x-axis from 2007 to 2012. Three conditions are compared: Mixed, Industrial, and Domestic, as indicated in the legend. All three conditions are stacked and compared in each year category. | The chart shows that the Cadmium discharge peaked in 2009 (around 2300 kg) and 2007 ( around 2240 kg), dominated by mixed conditions. Discharge is very low in both 2008 and 2010, with 0 levels in 2011. | Bar | Stacked Bar |
ACM_3542.png | This multiple line chart shows predicted Output on the y-axis in range of -1 to 3 the prediction output against the Testing samples on the x-axis in range of 0 to 300. Two conditions are compared as indicated in the legend: Real value, and Predicted value. | The chart reveals that the actual value of the output tends to be higher than the predicted value across almost all testing samples. | Line | Multiple Line |
0863.png | This horizontal box plots shows the TLX impression score (out of 20) on the x-axis across the 4 types of tools on the y-axis: baseline, Menue&beacon, Hints, nad Quadrant Zoom. | Quadrant Zoom has the lowest TLX Impression score, around 10. Other three tools have higher score, in the range of 12 to 15. | Boxplot | Simple Boxplot |
springer_4475.png | This bar chart compares the expected selling price and production cost for four conditions. The y-axis lists the four conditions: Protein CP food, Protein SP food, Protein CP feed and PRotein SP feed. The x-axis shows euro per kilogram ranging from 1 to 100. Two conditions are compared as indicated in the legened: expe... | The chart shows three conditions are shown to be profitable: the protein cP food, Protein SP food and Protein CP feed have a all an expected sellin price that is higher then the production cost. Protein SP food has the biggest difference between the two conditions. | Bar | Group Bar |
0441a.png | This bar chart compares Average Voews Per Query for Prediction and Standard. Average views per query is shwon on the y-axis ranging from 0 to 1.5, and the x-axis shows Prediction and Standard. | Standard shos higher average views per query around 1.25; Prediction shows lower average views per query, around 1.08. | Bar | Simple Bar |
springer_0735.png | This scatter plot with regression line is titled 'age at maturity - regions 1 & 2' and shows A50 (yr), ranging from 2.0 to 3.5, against year, ranging from 2009 to 2021 on the x-axis. There are two datapoints per year, 'functional' and 'biological'. | The chart reveals no clear pattern in A50 over time. the maximum is reached at year 2012, with functional and biological reaching 3.3 and 2.9 respectively. Overall, functional datapoints tend to have a higher A50 than the biological datapoint in the same year. | Scatter | Scatter plot with error bars |
ACM_2077 copy 2.png | This multiple line chart shows the number of worked-through problems (y-axis) ranging from 0 to 200 across different months (x-axis): October to April. The chart compares two countries: Germany and the Netherlands. | The chart shows that the Netherlands consistently outperforms Germany in terms of worker-threshold problems throughout the year. | Line | Multiple Line |
springer_4047.png | This grouped bar chart shows the relationship between cost and the hydrogen carrier. The x-axis lists six different hydrogen carriers, and the y-axis represents the cost in US dollars per 100 km, ranging from 0 to 1.2. The legend shows six different countries that are compared for each carrier. | The chart shows that the most expensive hydrogen carrier is Pipeliine - Onshore, followed by NH3, the least expensive Methanol. The country where it's most expensive ist Libya, followed by France and then th UAE, the least expensive is Saudi Arabia. | Bar | Group Bar |
ACM_1246 copy.png | This is a heatmap titled 'Semantic Similarity Threshold = 96.00'. The y-ayis display Max Cluster Size (ranging from 500 to 100) and Max Time Passed (in minutes) ranging from 52 to 92 on the x-axis. The color bar on the right represents the ARI score, with values ranging from 0.5 to 0.9. Each cell has a value represents... | This chart shows that when the max cluster size is 100, all the max times passed have the same ARI score at 0.54, and when the max cluster size is equal or above 200, all the max times passed have the same ARI score at 0.73. | Heatmap | Heatmap |
Springer_2nd_0457.png | The chart is a pie chart titled "ESG Scores," displaying three key components: Environmental (ENV), Social (SOC), and Governance (GOV) scores. The legend differentiates these components using colors and highlights specific metrics. The axis labels indicate various scores, while scores range from 0 to 35. | The chart shows varying ESG scores, with governance: management scoring highest at 35, while social elements are notably lower, emphasizing a trend towards governance priorities. Within the ENV score, resource use scores highest (20). | Pie | Pie |
ArXiv_3rd_709.png | This horizontal bar chart shows the value of mean(|SHAP value|)(average impact on model output magnitude) on the x-axis from 0.0 to 0.5, and 14 variables on the y-axis. | The chart shows that the "perimeter_worst" has the highest average SHAP value, exceeding 0.4. Followed by the features of 'perimeter_mean', 'texture_worst', etc. There are 6 features have almost 0 average SHAP value. | Bar | Simple Bar |
springer_4599 copy.png | This bar chart is titled "Relative Influence Plot". The x-axis shows the relative incluence, ranging from 0 to 40. The y-axis shows different variables. | The chart shows that only 5 variables have a relative incluence. The highest being at around 37 and the lowest at around 3. | Bar | Simple Bar |
1875.png | This bar chart, titled "Frequency Thresholds for Tactile 'Click' Sensations," shows the frequency values in Hertz for three body locations: Fingertip, Hand Dorsum, and Forearm. The y-axis represents frequency, ranging from 0 to 160 Hz, while the x-axis lists the body locations. Each bar represents the frequency thresho... | The frequencies are relatively similar across the three body locations, with the Forearm showing a slightly higher frequency threshold. | Bar | Simple Bar |
arXiv_1103.png | This bar chart shows the specificity (TNR) for various models. The y-axis represents specificity (TNR), ranging from 0 to 1, and the x-axis lists different machine learning models, including BioSerenity-E1, MTP-Virgin, EEGNet, DeepSOZ-max, TGCN, and TSD. | "The chart shows that the most significant factors are FC3NET and EEGWNET followed by MTP-Virgin and BioSerenity-E1. Other models have zero sensitivity. | Bar | Simple Bar |
ArXiv_3rd_652.png | This multiple line chart shows "number of ENVs" on the left y-axis from 0 to 10, 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. | As time resolution increases, the number of ENVs declines sharply from 8 to 2. The ENV₀ fraction remains near zero throughout. | Line | Multiple Line |
arXiv_1009.png | This multiple line chart is titled "Trajectories of the Cube Block" across different target positions on the x-axis, ranging from 0.3 to 0.46. The y-axis represents the position ranging from -0.03 to 0.1. The chart includes five lines representing different techniques and a Target Position. | The chart shows that the Trajectory in Simulation has the closest position to Target Position, while the 1st PPO Trajectory has the biggest deviation from the target position. | Line | Multiple Line |
ACM_2142.png | This multiple line chart shows the CDF against the RTT. The CDF is on the y-axis in range of 0 to 1, and the RTT (ms) is on the x-axis in range of 0 to 500. The chart compares four conditions: same city, different city, different country, and different continent. | The chart shows that the CDF increases with time for all four conditions, but the "same city" and "different country" increase exponentionally just after 0, while "different county" increases slowly, and"different continent" only starts increasing at 100ms and then increases direct. | Line | Multiple Line |
ACM_0005.png | This horizontal bar chart shows the number of papers published by the authors themselves, categorized by their academic degree. The y-axis lists the categories: K-12, Undergraduate, Graduate and Professional. The x-axis represents the number of papers, ranging from 0 to 120. | The chart shows that "Undergraduate" papers dominate the field with 112 instances, while "Graduate" papers are fewer, with 16 instances. "K-12" papers are minimal, with only 2 instance, same with "Professional" with only 1 instance. | Bar | Simple Bar |
springer_2575 copy 2.png | This scatter chart shows the "APOE" across different Cancer types. The x-axis represents the different Cancer types, while the y-axis represents the "APOE" ranging from -10 to 15. | The chart reveals that all the Cancer types remain between 0 and 15 APOE. However, the LAML type has a lower APOE reaching -2.5. The Cancer type with the highest APOE is LMC with 15. | Scatter | Simple Scatter |
arXiv_0965.png | This multiple line chart shows the real-sc outcome difference against the time in days. The y-axis shows the Real-SC Outcome Difference, ranging from -2 to 2, and the x-axis shows the Time in days, ranging from 2020/01/02 to 2020/02/21. There are two conditions: Control Diffs and Treated Diff. | The chart shows that the "Treated Diff" increases from 0 to 2 at 2020/01/27, while the "Control Diffs" have many lines varying all over the graph. | Line | Multiple Line |
Springer_2nd_0876.png | This is a line chart titled "Fishery Production in EU14 Countries," showing production in metric tons from 1990 to 2018. The x-axis represents the year, and the y-axis ranges from 78.5 to 83 metric tons. | Fishery production rose gradually to a peak around 1998, then declined steadily through 2010, reaching a low around 2014. A slight recovery followed, but levels remained below the earlier peak of 82 Tons. | Line | Simple Line |
ACM_4282.png | This grouped bar chart shows the number of initializations on the y-axis, from 1 to 10 power of 8; 4 datasets are compared on the x-axis. Six methods are compared for each dataset, as indicated in the legend. | The (epsilon, Delta r) minus APX method consistently has the highest initialization counts, especially on datasets Orkut and wise. In contrast, CPT-Dec and APX-Dec maintain low counts.performance. | Bar | Group Bar |
springer_2249 copy.png | This grouped bar chart shows the Iron concentration of mesencephalon in nanograms per gram on the y-axis in range of 0 to 40 against Days on the x-axis ranging from 14 to 90. 3 conditions are compared as seen in the legend: Control, 100 milligrams per kilogram ATR, and 200 milligrams per kilogram ATR. | The chart shows that the control condition has the lowest concentration after 90 days at about 20 nanograms per gram, followed by 100 milligrams at about 30 and 200 milligrams at about 31. | Bar | Group Bar |
ACM_2262.png | This multiple line chart shows the relationship between Loss Value and Epochs for two conditions: Train Loss and Train Accuracy. The x-axis represents Epochs, ranging from 0 to 200, while the y-axis on the left represents the Loss Value, ranging from 0.0000 to 0.0080. Then, the y-value on the right represents accuracy ... | The train accuracy increases with epochs, then stagnates around 0.95 accuracy rate. The train loss decreases loss value as Epochs increases. | Line | Multiple Line |
ACM_0251.png | This multiple line chart shows Accuracy on the y-axis in range of 0.0 to 1.0 against the academic week on the x-axis in range of 0 to 10. Three conditions are compared as indicated in the legend: Standardized, Curved, and Most Frequent Quessing. | The chart shows that the "Standardized" model outperforms both the "Curved" and "Most Frequent Quessing" models across all academic weeks, with accuracy consistently above 0.5. The Standardized Line ends at around 0.75, while the Curved condition reaches around 0.65. The Most Frequent Quessing Line stays stable at arou... | Line | Multiple Line |
springer_2838 copy 6.png | This line chart represents the irrigated area in square km on the y-axis ranging from 0 to 700 the irrigated area in % of Sub-basin Area is represented on the second y-axis, ranging from 0.0 to 8. The x-axis represents the years form 1990 to 2020. | The chart shows a general upward trend in the percentage of irrigated sub-basin area, with minimal fluctuations over the years and a notable spike in 2015. | Line | Simple Line |
ACM_2006.png | This bar chart shows the Soil Moisture (Volumetric Water Content) of 2 different compositions. The y-axis displays the Soil Moisture, ranging from 0% to 100%, and the x-axis displays 2 compositions: Leaf Mulch and Bark Mulch. | The chart shows a clear divergence between the two compositions, with the "Bark Mulch" exhibit significantly higher decomposed percentage at 13%, whilst the "Leaf Mulch" is lower at 7%. | Bar | Simple Bar |
ACM_2580.png | This line chart shows the "Loss" on the y-axis in range of 0.2 to 1.4 against "Iterations" on the x-axis in range of 0 to 70. Two lines are compared as indicated in the legend: "Umbalanced datasets" and "Balanced datasets". | The chart shows that the Unbalanced Dataset has a higher loss as the iteration count increases compared to the Balanced Dataset. | Line | Multiple Line |
0088b.png | This line chart, titled "Average Intensity," compares the intensity levels in dB for two conditions, "Calling" and "Not calling," as shown in the legend. The x-axis represents time in milliseconds (ms), ranging from 0 to 700, while the y-axis displays intensity in decibels (dB), spanning from 0 to 70. | The chart reveals that intensity starts at a high level in both conditions, with "Calling" consistently higher than "Not calling" throughout. Both lines show a decreasing trend, with intensity dropping steadily from the beginning to the end of the time period. The "Calling" condition maintains a larger intensity gap ov... | Line | Simple Line |
extra2_0143.png | The chart is a box plot titled "Mn." It compares manganese (Mn) levels (mg/kg) in two categories: Root and Shoot. The y-axis is labeled "Mn (mg/kg)" with a scale from 0 to 400. The median and quartiles are indicated by the boxes, while the whiskers represent the range of the data. Asterisks denote statistical significa... | The chart shows that Mn levels are significantly higher in the Shoot (200) compared to the Root (approximately 100), indicating a notable relationship between plant part and manganese concentration. This trend suggests varying nutrient uptake efficiency by different plant structures. | Boxplot | Simple Boxplot |
springer_2600.png | This heatmap chart shows the predicted number of vehicles. The x-axis lists different types of True vehicles, including cars, buses, trucks, and motorcycles. The y-axis represents the Predicted types of vehicles. Each area is labeled with a value marking the threshold from 0 to 0.8 on the y-axis. | The chart reveals that all Predicted vehicles that are true have a value of around 0.8, while anything outside is either 0 or has a value below 0.1. | Heatmap | Heatmap |
springer_2757.png | This simple line chart shows the PV & Load Power (W) from 0 to 1200 on the y-axis, while the x-axis shows time (s) in range of 0 to 6. Two lines are indicated in the legend: blue line for Ppv and red line for Pload. | The Pload condition shows a notable fast increase at 0 seconds, reaching 700 W. After that, the Pload condition is stable with minimal, barley notable fluctuations over time. The Ppv condition shows an increase at 0 seconds as well, reaching 400 W. After that, it shows a general increase as time increases. The Ppv cond... | Line | Multiple Line |
springer_1960.png | This boxplot shows the Fisher's Z Transformed Correlation Coefficient internet gaming disorder value with 95% confidence interval for many different studies. The x-axis shows the z value ranging from -0.2 to 1.2, and the y-axis lists the studies. | The boxplot shows that Monacis et al. has the highest z value of 1.01. The lowest belongs to Chen et al., at 0.06. | Boxplot | Simple Boxplot |
arXiv_0409.png | This multiple line chart is titled "hospitalizations" in DRAM. The x-axis represents the date from 04/03 to 06/02, and the y-axis shows the number between -50 and 75. Three Pushforward Intervals are shown, along with a line representing the Observed values. | The observed values fluctuate while going down over time from around 25 on 04/03 to around -10 on 06/04. | Line | Multiple Line |
arXiv_0916.png | This grouped bar chart is titled "Sexual Orientation: Heterosexual" and shows the share of comfortable with gay individuals across different co-workers. The y-axis represents the share of comfortable with gay individuals, ranging from 0.4 to 1, while the x-axis lists various co-workers: Double List Supervisors, Direct ... | The chart shows that the "Direct Question" has the highest share of comfortable gay individuals, around 0.85, followed by "Double List" around 0.65, however, "Supervisors" have a slightly lower share than the other two. | Bar | Group Bar |
springer_1564 copy.png | This heatmap displays correlation coefficients between bacterial groups (rows) and physicochemical properties (columns), with color intensity on the right part of the y-axis representing the strength of the correlation (ranging from 0.00 to 0.50). Red shades indicate stronger positive correlations. | “Shared” bacteria show high correlation with AP, As, and RAP. HYS-Enriched group is strongly associated with As, while BPS-Depleted bacteria correlate most with RAP and TN. | Heatmap | Heatmap |
ACM_3495 copy.png | This multiple line shows the angle error as a function of time for four different configurations: 2 km, 3 km, 4 km, and 5 km. The x-axis represents the angle, ranging from 0 to 180 degrees, and the y-axis represents the angle error, ranging from 1.95 to 2.2. | The plot shows that as the angle increases, the error decreases across all conditions up to around 80 degrees, after which the error begins to decrease. A regression line highlights this trend for each condition. Among the conditions, the 2 km setting exhibits the highest error, followed by the 3 km, 4 km, and 5 km con... | Line | Multiple Line |
ACM_1777.png | This grouped bar chart is titled "Chi Squared Test", with 5 datasets represented by the x-axis: A, B1, B2, C, D. The y-axis shows a scale in range of 0.0 to 0.8. 4 generators are compared as indicated by the legend: CTGAN, GPT2, DialoGPT, DistilGPT2 | The chart shows dataset B1, B2 C, and D have the highest value with generator CTGAN. GPT2 has the lowest value of 0 with dataset D, while its the highest value for dataset A with around 0.75. | Bar | Group Bar |
0876a.png | This scatter plot with error bars is titled "Endpoint Study". It has Score in range of 0 to 4 on the y-axis, 3 metircs shown on the top x-axis: Hand Match, Tolerability, and Ease. 2 conditions are compared as seen in the legend: MPC-E, nad HR. | MPC-E WarpType shows higher Scoren than HR WarpType among all three metrics. | Scatter | Scatter plot with error bars |
arXiv_0656.png | This multiple line chart shows the relationship between the relative time (x-axis) and the GPU usage (y-axis) for four conditions. The x-axis represents time, ranging from -1m to +1m, and the y-axis represents GPU usage, ranging from 14.0 to 15.0. | The actual usage (red line) consistently trends above all three prediction models. Among the models, 1D-CNN and FC-LSTM align more closely with the true signal, while GRU (orange line) tends to underpredict and lag behind. | Line | Multiple Line |
springer_2265 copy.png | This bar chart titled Astronomy Publications from India shows the peer reviewed astronomy publications sourced from ADS on the y-axis ranging from 0 to 2500, and the year on the x-axis from 1998 to 2022. | The number of publications per year increases with the years. The year with the highest number of publications was 2021 at about 2200. The year with the lowest number of publications was 1997 at about 250. | Bar | Simple Bar |
springer_3797.png | This pie chart shows the percentage breakdown of different topics. The chart segments are labeled with 11 categories, including Economics, Environemnts Science, Social Science, Busincess, Management, Energy, Computer science, Engineering, Mathematics, Decision Science, Multidiscipliary, and other. Each category is repr... | The largest segment, Economics, dominates at 21%. The second largest, Environmental S., follows at 16.8%. The smallest segments are Multidisciplinary and Decision Science, both at 2%. | Pie | Pie |
ACM_1211.png | This is a multiple line graph titled "F1 Score vs Epochs (EMNIST)". The x-axis represents the epochs (from 0 to 25), and the y-axis represents the F1 score (ranging from 0.78 to 0.88). The The legend identifies six models with different color-coded line with marker. | The models show a general increase in performance as the number of epochs grows, with KANICE reaching the highest F1 score (around 0.88) at approximately 25 epochs. CNN shows the lowest F1 score, stabilising at around 0.85.. | Line | Multiple Line |
springer_1430.png | This grouped bar chart shows conditions: PFF, RFF and population density across regions. The x-axis lists regions, while the left y-axis (0–60) indicates the number of events, and the right y-axis (0–9000) represents population density. | elangor and Kuala Lumpur show high population density, while Selangor has high PFF/RFF events. Regions like Hulu Selangor and Sepang also show elevated activity. | Bar | Stacked Bar |
Springer_2nd_0037.png | The chart is a scatter plot showing Number of Photos Taken versus the SBSOD Score. The x-axis represents SBSOD scores (ranging from 2 to 6), while the y-axis shows the number of photos taken (from 0 to 15). The legend includes a fitted regression line with a shaded confidence interval. | The chart illustrates a negative correlation (R = -0.56, p = 0.045), indicating that as SBSOD scores increase, the number of photos taken tends to decrease. | Scatter | Scatter plot with regression line |
springer_4214 copy.png | This boxplot shows the Mean Absolute Error for four different methods. The y-axis represents the Mean Absolute Error in range of 10 to the power of -3 to a little over 10 to the power of -1. The x-axis shows the four methods: MARS, MLP, SVM and DT. | The plot shows that the SVM method has the highest Mean Absolute Error score, which means it's the least accurate. The DT method has the lowest Mean Absolute Error score, which means it's the most accurate, followed by MARS. | Boxplot | Simple Boxplot |
extra_0339.png | This pie chart is titled 'Relative abundance of microplastic shape in surface Sediment'. It lists four categories in the legend: 'Fragment', 'Film', 'Granule', and 'Fiber'. Each category is represented by one segment in the pie, with the percentage labelled. | The chart reveals that the Fragments constitute the majority at 44.8%, Film takes up 41.2%, followed by the section of Granule, with 9.9%. Fibers are the least abundant at 4.1%. | Pie | Pie |
arXiv_0408.png | This multiple line chart is titled "CRPS" in hospitalizations. The x-axis represents the date from 04/03 to 06/02, and the y-axis shows the number between 0 and 200. Two conditions are compared: DRAM and SVI. | The chart shows that DRAM has a mean of 31.87, while SVI has a mean of 31.59. The two lines follow each other closely, having peaks and dips in the same places. The highest peaks are at about 120 for both lines around 05/10. | Line | Multiple Line |
arXiv_0953.png | This multiple line chart shows the average computer usage against the time in days. The y-axis shows the Average Computer Usage, ranging from 8 to 13, and the x-axis shows the Time in days, ranging from 2020/03/16 to 2020/04/24. There are two conditions: Synthetic Control and Treatment. | The chart shows that both conditions have similar average computer usage, ranging from 9 to 11, with a trough at 2020/04/07. The Synthetic Control condition generally shows a higher average computer usage than the Treatment condition. | Line | Multiple Line |
2769c.png | This boxplot shows Perceived Arousal across Two Durations. The y-axis shows Arousal in range from 1 to 7, the x-axis lists two types of duration. | Short duration has lower arousal, around 2.3, the long duration has higher arousal, around 3. | Boxplot | Simple Boxplot |
1649d.png | This box plot shows the precision metric for user annotations on the "Immigration" and "Abortion" topics, separated by the article source leaning (Left and Right). The y-axis represents precision, ranging from 0 to 1. The data shows that precision is generally lower across all categories, with "Abortion" (Right) showin... | The plot indicates that "Abort [R]" has the highest median precision, demonstrating a relatively higher rate of accurate annotations on abortion-related content among right-leaning participants. "Immig [R]" shows moderate precision with a wide spread, while "Immig [L]" and "Abort [L]" have lower medians, indicating low... | Boxplot | Simple Boxplot |
ACM_1781.png | This scatter plot is titled 'Mean Silhouette for s-clusters'. The x-axis shows Number of clusters (K) in range of 2 to 10, the y-axis shows Mean Silhouette in range of 0.2 to 0.4. | The plot shows a negative trend between the number of clusters and the mean silhouette. K-value 5 is marked as having value 0.26 around the elbow point. | Scatter | Simple Scatter |
springer_4376.png | This grouped bar chart is titled "Patients with Positive Experience with the Infomred Consent Process". The x-axis shows fivecategories: Very satisfied/satisfied, IC given by surgeon, >24 hours before operation, Enough Time, Family support. The y-axis shows the percentage, ranging from 0.0% to 100.0%. There are two con... | The chart reveals that highest percentage for the Elective condition has the category Enough Time with 94.2%, followes by Very satisfied/satisfied with 90.4%. For the Emergency condition the highest percentage has Very satisfied/satisfied with 83.7%, followed by >24 hours before operation with 76.9%. | Bar | Group Bar |
ACM_0155.png | This grouped bar chart has no title. The y-axis shows the percentage ranging from 0 to 100%. The x-axis shows different categories named: OS-G2, OS-G4, OS-P, DB-G2, DB-P, DB-A, CS1-P and CS1-A. The bars are also divided into 6 different categories: Late Grace, On-time Grace, Late Penalty, On-time Penalty, Late Auto and... | The chart shows that the percentage of late penalties is generally lower across all conditions compared to the percentage of on-time penalties. This also applies to the Late Grace and On-time Grace category. In the category DB-A the late auto percentage is higher than the on-time auto. | Bar | Group Bar |
ACM_2906 copy 2.png | This multiple line chart titled HH Polarization shows the relationship between "Observation duration" on the x-axis (ranging from 7 to 12) and "Detection Probability" on the y-axis (ranging from 0.1 to 1.0). The chart includes five lines representing different heuristic detection methods: "Three-Feature Detector," "Fra... | The plot shows that all models exhibit a positive trend, with the detection probability increasing as the observation duration increases. The models with higher detection probability are mentioned herein, LOF Detector and iForest Detector. | Line | Multiple Line |
1669b.png | This figure contains four scatter plots showing the absolute error of orientation (in degrees) across different marks (Circle, Tri-Up, Tri-0, Tri-20, and Tri-40) for four residual bandwidths (0.05, 0.1, 0.15, and 0.2). Each plot shows five CI error bars for the different marks. | Across all residual bandwidths, the estimated absolute errors are similar for the different marks. However, the error bars show greater variance as the residual bandwidth increases, particularly at 0.2, where the errors are the highest, exceeding 12 degrees. | Scatter | Simple Scatter |
extra2_0278.png | The chart is a scatter plot with error bars. The x-axis is labeled "School grade" and ranges from 7th to 9th, while the y-axis is labeled "Beta Coefficients" and spans from -1.0 to 1.5. | The chart shows that 7th grade has a beta coefficient of -0.66, while 8th grade peaks at 0.93. The 9th grade is at 0, revealing significant variability. | Scatter | Scatter plot with error bars |
extra_0481.png | This heatmap has predicted classes on the x-axis and true classes on the y-axis. Both axes have two categories of labels: “High” and “Low” classes. Cell values represent counts, and a color bar that scales from 0 to 6000 is shown on the right side. | The plot shows that classification is highly accurate for the "High" class (6834 correctly classified and 426 wrongly classified), but shows reduced accuracy for "Low", with only 6994 correct and 269 misclassified. | Heatmap | Heatmap |
3444.png | This pie chart shows Number of Records per Research Direction. 4 Tyeps of research directions are compared: AI for XR, Intelligent VAs, XR and AI applied, and XR for AI. Each resaech direction is represented by one slice of the pie. | Ai for XR has 60.1% numebr of records, Intelligent VAs has 13.8% of records, XR and AI applied has 23.8% of records, and XR for Ai has only 2.3% of records. | Pie | Pie |
springer_2221 copy 4.png | This bar chart shows the The Cleaved Caspase-1/beta-actin of three conditions. The x-axis lists the conditions: Control, Model and HPTQ, and the y-axis shows the Cleaved Caspase-1/beta-actin ranging from 0.0 to 1.0. | The control condition has the lowest y-value of about 0.4, while the model condition has the highest around 0.8. HPTQ is around 0.5. | Bar | Simple Bar |
0547.png | This bar graph shows the average scores on the Y-axis in range of 0 to 1.0, against the OI and BI conditions on the x-axis. | Participants scored higher in the OI condition with an average score of 0.58 compared to the BI condition with an average score of 0.54. | Bar | Simple Bar |
arXiv_0492.png | This multiple line chart shows Energy Consumption Before and After Optimization. The x-axis represents time as dates from january first 2024 to february eighth 2024, the y-axis shows energy consumption in kilowatt hours. Two conditions are compared as indicated in the legend: Before Optimization, and After Optimization... | The chart reveals that both conditions exhibit a similar trend, but the "After Optimization" group is consistently lower than the "Before Optimization" group in terms of energy consumption. | Line | Multiple Line |
ArXiv_3rd_195.png | This heatmap is titled 'Proposed Model Confusion Matrix'. The y-axis shows True labels, and the x-axis shows Predicted Labels. Both axes have 3 categories: 'Covid', 'Normal', and 'Pneumonia'. The cell values indicate classification counts. | The chart shows that correct classifications are the most cases: 239 COVID, 238 normal, and 242 pneumonia. Misclassifications are minimal, with only a few normal and pneumonia samples confused with other classes. | Heatmap | Heatmap |
extra_0658.png | The chart is a box plot titled "FKGL." It includes three categories: ChatGPT, Gemini, and Copilot. The x-axis shows the three models, while the y-axis indicates the FKGL values, ranging from 4 to 15. | The box plot reveals that ChatGPT has the highest median FKGL, while Copilot shows considerable variability. A perceptual trend indicates that Gemini performs closely to Copilot, suggesting similar comprehension levels, while ChatGPT's performance stands out significantly across the dataset. | boxplot | Simple Boxplot |
extra2_0081.png | The chart is a box plot. It compares two categories: MSW and waste generation. The y-axis is labeled "waste generation (kg/c/d)" and ranges from 0.1 to 1, while the x-axis represents the two categories. | The box plot reveals a higher median waste generation for MSW compared to regular waste, with a slight difference in the interquartile ranges. This suggests greater variability in MSW. | Boxplot | Simple Boxplot |
arXiv_0596.png | This is a multiple line graph titled “Overall Survival by Cluster.” The x-axis shows overall survival in years, and the y-axis shows the proportion of patients surviving. The legend distinguishes five clusters—C1 (mustard), C2 (green), C3 (light blue), C4 (navy), and C5 (purple). | The graph compares patient survival over time across five clusters. Cluster C3 shows the best survival rate, while C5 and C1 show the poorest outcomes. This highlights survival variability between patient subgroups. | Line | Multiple Line |
springer_1564 copy 8.png | This boxplot illustrates Cd (mg/kg) concentrations across CK, PBRS, and HuaYS treatments. The y-axis ranges from 0.3 to 0.6. | HuaYS shows the highest median cadmium content with a relatively narrow spread, followed by PBRS with a broader distribution. CK has the lowest values with minimal variability. The significance letters (a, b, c) suggest statistically distinct differences in cadmium accumulation across all treatments. | Boxplot | Simple Boxplot |
extra_0379.png | This grouped boxplot shows the value of Prosociality Stress on the y-axis from 1 to 5, against three group categories listed on the x-axis: Passive, Intervention, and Waitlist. Two Time conditions are compared within each group, as indicated in the legend: Time 1 and Time 2. | This chart shows that the Time 1 and Time 2 conditions have different scores in Prosociality only in the Intervention group, where the Time 2 condition has a higher median score (4.0) than the Time 1 condition (3.8). Both Time 1 and Time 2 conditions have the same median score of Prosociality in both groups of Passive ... | Boxplot | Grouped Boxplot |
pie_160.png | The chart is a pie chart titled "Most Disapprove of Government’s Recovery Efforts." It includes a legend with categories: Disapprove, Approve, and DK. The chart uses color to encode data, with percentages labeled directly on each segment.Each category is represetend by one slice of the pie. | The majority, 60%, disapprove of the government's recovery efforts, while 37% approve. A small portion, 4%, are undecided (DK). | Pie | Pie |
Springer_2nd_0848.png | This is a heatmap titled "HCHO 2019." The map spans from 5°E to 6°30′E (longitude) and 6°N to 7°30′N (latitude). The legend classifies formaldehyde (HCHO) levels into green for LOW, yellow for MEDIUM, and red for HIGH. | The map displays spatial variations in formaldehyde concentrations across a region in 2019, showing higher levels (red) clustered in central and southeastern zones, with lower levels (green) more in the west. | Heatmap | Heatmap |
springer_0752.png | This line chart shows the number of geometric figures visited by a participant. The x-axis lists various figures, and the y-axis shows visits, ranging from 0 to 4. one line represents an individual (Viktor), one line represents the adult mean. | The chart shows that Viktor has a higher number of visits than the average for all figures, especially for the similar prototype rectangle and similar rotated rectangle. | Line | Multiple Line |
arXiv_0577.png | This is a dual-line graph titled “Estimates of Death Rates of Unaware Infections.” The x-axis shows dates from 9/1/2020 to 3/8/2022; the y-axis ranges from 0.0006 to 0.0022. The legend distinguishes the Omicron variant (orange) and the Wild-type strain (blue). | The graph shows the wild-type strain had higher and more variable death rates early on, while the Omicron variant maintained lower, relatively stable rates. This suggests reduced fatality during the Omicron period. | Line | Multiple Line |
0432.png | The line chart shows is titled '(a) Accuracy Curves", it presents Training Accuracy against Epochs for two conditions: Sensor+Screenshot(random labels) and Sensor+Screenshot (true labels). Training Accuracy is shown on the y-axis from 0.5 to 1.0, 21 epochs is shown on the x-axis. | In the condition of Snesor+Screenshot(random labels), trainig accuracy stays consistent around 0.55 through 21 epochs; while sensor+screenshot(true labels) shows increasing trend through 21 epochs. | Line | Multiple Line |
ArXiv_3rd_253.png | This multiple line chart shows the accuracy on the y-axis in the range of 20 to 100 against the step on the x-axis in the range of 1 to 8. Three conditions are compared as indicated in the legend: Angles analysis, Construction forces analysis, and Neutral stimulus analysis. | The plot shows that the accuracy grows as steps increase for all three conditions. The "Construction forces analysis" method has the highest accuracy across all steps. The 'Angles analysis' and 'Neural Stimuli analysis' have similar and lower accuracy across all steps. | Line | Multiple Line |
springer_1775 copy 2.png | This scatter plot compares predicted and true energy values, with both axes ranging from -1.0 to 1.0. The red dashed regression line indicates perfect predictions. Density on the right side side of the y-axis ranges from 0.0 t 1.0. | Most data points are tightly aligned along this line, forming a high-density diagonal cluster (indicated by the green-yellow shades on the color bar). The strong alignment suggests very high model accuracy and low prediction error across the observed range. | Scatter | Scatter plot with regression line |
springer_0204 copy.png | This multiple line chart is titled "SSI". It shows the number of applications on the y-axis from 70'000 to 100'000. The x-axis shows six numbers (probably years and months) 2011/9, 2013/5, 2015/1, 2016/9, 2018/5 and 2020/1. Three conditions are being compared: Actual, Spatial and AR(2). | The chart shows a similar trend for all conditions. They start around 100'000 and decrease to 80'000. All lines are fluctuating, and show quite the same results but the actual condition shows slightly lower values then the other two. | Line | Multiple Line |
extra2_0306.png | The chart is a box plot showing "MSWS-12" for two groups: intervention and control. Per group, two boxplots are shown: Baseline and 12 weeks. the x-axis shows the four boxplots, the y-axis ranges from 0 to 30 and is labeled "MSWS-12". | The box plots reveal that the intervention group showed a significant reduction (indicated by an asterisk) in MSWS-12 scores at 12 weeks compared to baseline. The median value changes from 45 to 35 approximately after 12 weeks. The change in MSWS-12 score is also visible in the control group, where the median values ch... | Boxplot | Grouped Boxplot |
Springer_2nd_0488.png | The chart is a raster histogram titled "Raster Histogram." The legend indicates that the plot shows Band 1. It employs frequency as the y-axis label and pixel value as the x-axis label, both on linear scales. | The histogram reveals a significant peak at pixel values around 0.4, showing that a large number of pixels cluster here. For pixel values greater then 0.4, the frequency tends to be higher than for lower pixel values. This trend suggests a concentration of pixel values at and beyond 0.4. The highest frequency is for pi... | Bar | Simple Bar |
ACM_1208.png | This is a multiple-line graph title, F1 Score Vs Epochs (MNITS). The x-axis represents the epochs (ranging from 0 to 25), and the y-axis represents the F1 Score (ranging from 0.955 to 0.995). The legend identifies different six models, each represented by a line with marker of varying colours. | Most models show an increase in F1 score over epochs, stabilising at a high value after around 15 epochs. The KANICE model performs slightly better than the others, consistently reaching the highest F1 score, while CNN stay behind other models in performance. | Line | Multiple Line |
0025.png | This bar graph represents movement variability for touch input tasks across different drag angles. The x-axis shows Drag Angles in the range from 0 to 330. and the y-axis shows movement veriability(mm), ranging from 0.0 to 3.0. | The lowest variability is found at 0, 90, 180, and 270 degrees, with only 2.0 mm movement variability. 0, 150, 210, and 330 degrees have highest movement variability, around 3.0mm. | Bar | Simple Bar |
arXiv_0136.png | This grouped bar chart is titled 'Female' and shows different medicines on the x axis. These include: Apixaban, Paracetamol, Dabigatran among others. The y-axis is unlabeled and ranges from 0 to 240. For each medicine two bars are shown, 'age' and 'prescribed'. | The chart reveals that WarfarinSodium and Apixaban have the highest y-value of around 210. | Bar | Group Bar |
ArXiv_3rd_183.png | This multiple line chart titled “3p3 example” plots SE (y-axis ranging from -2 to 0) against Time (x-axis ranging from -6 to 5). Three conditions are compared in the legend: 'Past 3 years', 'tru future', and 'pre future'. | The plot shows that the SE value for the 'past 3 years' condition is constant at 0 and then decreases to -2 during the time interval 1 to 3. The predicted and true future conditions are closely aligned, both of their SE values decreased from -2 to -6 during time interval 2 to 5. | Line | Multiple Line |
springer_4357.png | This scatter chart shows the relationship between different categories on the y-axis and on the x-axis the "SHAP value (impact on model output)", ranging from -0.3 to 0.2. | The chart reveals that for Surfacelocation the SHAP value is spread the widest in both directions. On the negative side of the SHAP value the Feature value is very high. | Scatter | Simple Scatter |
extra_0489.png | This pie chart shows the distribution of three categories: Water, Land and Air. | Air accounts for the largest share, with 59%. The Land follows with 25%, and Water has the least percentage, of 16%. | Pie | Pie |
springer_2451 copy.png | This boxplot shows the horizontal distance for different models. The y-axis shows the Horizontal distance in micrometers, ranging from 500 to 1500. The x-axis represents different models: Amud 1, Amud 2, Manot, Quneitra, and Qafzeh. For each model there are three aoi conditions: 1, 2, and 3. | The plot reveals that the "Qafzeh" angle has the highest median and variability, while "Amud 2" has the lowest median and variability. | Boxplot | Grouped Boxplot |
ACM_3495.png | This multiple line chart shows the range error in percent on the y-axis in the range of 10 to 10,09 against the angle on the x-axis in the range of 0 to 180. Four conditions are compared as indicated in the legend: 2Km, 3Km, 4Km and 5km | The plot shows that as the angle increases, the range error decreases across all distances. The 2 km condition consistently has the highest range error across all angles. All distance conditions including 2 km, 3 km, and 4 km exhibit a peak in range error around 90 degrees, followed by a gradual decline. | Line | Multiple Line |
springer_0657.png | This multiple line chart is titled 'CEC2017 UniModel F1' and plots Diversity measurement, ranging from o to 30 on the y-axis, against iteration number on the x-axis, ranging from 0 to 500. Four lines are plotted, representing different dimensions: 10, 30, 50, 100. | The chart shows similar behavior for the dimensions, but the higher dimensions have a higher diversity measurement for the same iteration. number. This diversity measurement decreases to 0 as the iteration number increases, with a sharp decrease for all lines at iteration number 250. | Line | Multiple Line |
ACM_1221 copy.png | This is a pie chart titled "Model trained on trajectories filtered for Diversity". The chart represents the distribution of three categories: A-INT, Diversity, and Readability. | The chart shows the following distribution of categories: “Diversity” makes up the largest portion, with 80 entries. “A-INT” accounts for 6 entries. “Readability” represents the smallest portion, with 14 entries. | Pie | Pie |
pie_27.png | The chart is a pie chart titled "Remote worker share during COVID-19 Russia 2020, by implementation degree." It includes a legend with categories: 1-25% employees, 26-50% employees, 51-75% employees, 76-100% employees, and "Difficult to say."Each implementation degree is represetned by one slice of the pie. | The largest segment, 76-100% employees, is 35%. The smallest, "Difficult to say," is 1%. The chart shows a varied distribution of remote work implementation during COVID-19 in Russia. | Pie | Pie |
0125a.png | This is a horizontal stacked bar chart titled "Manual edits between runs?" comparing two systems: Sandbox and Chaining. The x-axis represents the percentage (from 0% to 100%) of manual edits between consecutive runs, while the y-axis lists the two systems being compared. The bars are split into two categories: "Consecu... | The chart indicates that in the Sandbox system,around 50% consective runs and edited runs. In contrast, the Chaining system had a much higher proportion of Edit runs, accounting for approximately 70%, with only about 30% rconsective runs. | Bar | Stacked Bar |
0245.png | This bar chart shows the distribution of participants' responses to the statement "The code-reading questions have helped me learn the material in the class" using a Likert scale (x-axis). The y-axis represents the "Number of participants," ranging from 0 to 20. The Likert response options range from 1 to 7. | The most common response is a Likert score of 5, around 15 participants responded. The least common response is likertscore of 1, with 5 rparticipants responded. | Bar | Simple Bar |
springer_4361 copy.png | This scatter plot shows the relationship between the two axes: Axis.2 (14.7%) and Axis.1 (22.5%). The x-axis in range of -0.2 to 0.4 against the logit on the y-axis in range of -0.4 to 0.2. Three conditions are compared as indicated in the legend: T1, T2, and T3. | The chart shows that the timepoint T2 overall reaches the highest scores, followed by the timepoint T2. All timepoints are widely spread. | Scatter | Simple Scatter |
ArXiv_3rd_38.png | This bar chart is titled 'Ranked general population responses on how attributes might impact AI behavior for Africans'. The x-axis lists 20 attributes, and the y-axis indicates the percentage of respondents, ranging from 0% to 16%. | The graph shows that the category 'Colonial history of country of origin' has the highest percentage of respondents, around 16%, and the category 'Age' has the lowest percentage of respondents, only 4%. | Bar | Simple Bar |
arXiv_0162.png | This heatmap is titled 'Consistency map (Downsampled) 874mV'. It shows time in seconds on the x axis, ranging from 500 to 4000 seconds. Repeat number is shown on the y-axis, ranging from 0 to 20. The heatmap shows voltage for each combination of x and y. | The map shows a voltage close to 0 for all values, except at time instances just after 1000 sec, 1500 sec, 2000 sec, 2500 sec, 3000 sec and 3500 sec. Then, the voltage increases briefly to 0.20 V, for all repeat numbers, and returns to the basis level of apprximately 0 Volt after. | Heatmap | Heatmap |
springer_3345.png | This grouped bar chart shows the Time on the y-axis from 0 to 5 seconds, across 5 types of voice datasets on the x-axis: TCFA, OCFA, WOA, ACO and WO-ANT. | The chart shows that ACO dataset has the highest time, above 4 seconds. The WOA voice dataset has the lowest time, around 1 second. | Bar | Simple Bar |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.