Upload README.md with huggingface_hub

README.md

@@ -1,5 +1,5 @@
 ---
-license: apache-2.0
+license: mit
 task_categories:
 - image-feature-extraction
 tags:
@@ -13,16 +13,22 @@ size_categories:
 
 # CC12M-Camera
 
+![camera map collage](analysis/cc12m_camera_map_collage.png)
+
 Per-image **camera parameter** annotations for the **CC12M** (Conceptual 12M)
 dataset (~10.97M images across 2,176 shards; **10,902,847** valid annotations),
 predicted by the **Puffin** camera-centric multimodal model
 (Qwen2.5-7B + C-RADIOv3-H visual encoder).
 
+The collage above visualizes the predicted camera maps on sample images — each
+pair shows the **up field** (green arrows: the projected gravity-up direction)
+and the **latitude field** (colored contours: angle above/below the horizon).
+
 ## Format
 
-The archive mirrors the source CC12M WebDataset layout: one `.tar` per source
-shard (`0000.tar` … `2175.tar`), each containing one `.json` per image whose
-name matches the source image stem. (CC12M has no train/val split.)
+One `.tar` per source shard (`0000.tar` … `2175.tar`), each containing one
+`.json` per image whose name matches the source image key (shard-local index).
+Source images: `pixparse/cc12m-wds` (same 2,176-shard layout / keys).
 
 Each JSON holds the predicted monocular camera parameters:
 
@@ -43,7 +49,7 @@ Example:
 ## Camera-parameter distributions
 
 Histograms of the predicted roll / pitch / vertical-FoV over the whole dataset
-(proportion of valid samples per 10° bin; `parse_ok=False` samples excluded).
+(proportion of valid samples per 10° bin; `parse_ok=False` excluded).
 
 ![cc12m camera stats](analysis/cc12m_camera_stats.png)
 
@@ -51,18 +57,11 @@ Histograms of the predicted roll / pitch / vertical-FoV over the whole dataset
 |-------|------------------|-------------------|-----------------|
 | all (10.90M) | 0.2° / 0.0° / 5.7° | −2.8° / −0.3° / 11.8° | 32.3° / 29.0° / 10.6° |
 
-**Reading the distributions**
-
-- **Roll** is sharply peaked at 0° — web images are overwhelmingly shot
-  upright/level (even tighter than ImageNet, σ ≈ 5.7°).
-- **Pitch** is centered close to 0° with a slight downward bias (median ≈ −0.3°,
-  mean ≈ −2.8°) and a moderate spread.
-- **FoV** is concentrated in the **20–45°** range (median ≈ 29°), typical of
-  everyday photography, with a long tail toward wide-angle.
+- **Roll** is sharply peaked at 0° (web images shot upright/level).
+- **Pitch** is centered close to 0° with a slight downward bias.
+- **FoV** concentrates in 20–45° (median ≈ 29°), with a long wide-angle tail.
 
 ## Notes
 
 - Annotations are model predictions, not ground truth.
 - Generated with greedy decoding (deterministic) at 640px input.
-- Figure reproduces `scripts/annotation/stat_camera_captions.py` from the Puffin
-  repo (read directly from the packed tars).