Simplify dataset card and focus Viewer on usable inputs

README.md

@@ -30,87 +30,49 @@ configs:
 
 # PROWBench Dataset
 
-**Do Video Models Render What the Program Specifies?**
+PROWBench evaluates whether video models follow program-specified motion and interactions. It contains **170 episodes, 200 camera views, and 600 proxy videos**, with prompts, recorded world states, and RGB first frames where available.
 
-**TL;DR —** PROWBench evaluates whether video models render program-specified motion and interactions, using recorded world states and timestamped events as references. This **private input release** preserves the `episodes/` directory hierarchy with **170 evaluation cases, 200 selected camera views, and 600 proxy videos**. Dataset Viewer indexes all **600 representation samples** across a full split and five group splits, with one image preview per case. Depth maps and generated model results are not included.
+[Paper](https://arxiv.org/abs/2610.02205) · [Project](https://alaya-lab.github.io/PROWBench/) · [Gallery](https://alaya-lab.github.io/PROWBench/gallery.html) · [Code](https://github.com/Brian90709/PROWBench) · [Evaluation guide](https://github.com/Brian90709/PROWBench/blob/main/eval/USAGE.md)
 
-[Paper](https://arxiv.org/abs/2610.02205) · [Project page](https://alaya-lab.github.io/PROWBench/) · [Gallery](https://alaya-lab.github.io/PROWBench/gallery.html) · [Code](https://github.com/AlayaLab/PROWBench)
-
-## Overview
-
-PROWBench asks whether a generated clip depicts a prescribed, executable world evolution—not just a plausible scene. Its records include entity and camera trajectories, articulated states, and timestamped events, including events outside the camera's field of view. The same saved states produce synchronized proxy videos for selected views; text prompts describe the scene and action timeline.
-
-Each selected view has three input representations:
-
-| Directory name | Paper terminology | What it conveys |
-|---|---|---|
-| `white` | Coarse 3D | Scene shape, silhouettes, and articulated motion with neutral materials |
-| `mixed` | CWM proxy / semantic proxy | Semantically colored environment and simplified entities |
-| `colored-obb` | Colored-OBB | Oriented entity boxes with stable identity colors and directional cues |
-
-These videos are **conditioning inputs**, not RGB ground-truth videos or generated model results. The three representations of a view share its recorded camera and world evolution.
-
-## Benchmark settings
-
-| Paper setting | Directory under `episodes/` | Viewer split | Cases | Views | Proxy videos | Available RGB first frames |
-|---|---|---|---:|---:|---:|---:|
-| Main: Verified-FF, 5 s | `verified_first_frame/` | `verified_first_frame` | 40 | 40 | 120 | 40 verified appearance anchors |
-| Main: Unverified-FF, 5 s | `unverified_first_frame/` | `unverified_first_frame` | 90 | 90 | 270 | 0 prescribed anchors |
-| Challenge: long horizon, 30 s | `long_video/` | `long_video` | 10 | 10 | 30 | 0 |
-| Challenge: synchronized multi-view, 5 s | `multi_view/` | `multi_view` | 10 | 40 | 120 | 0 |
-| Additional domain-specific Gunfight cases, 5 s | `gunfight/` | `neon` | 20 | 20 | 60 | 20 available source first frames |
-| **Total** | | `full` | **170** | **200** | **600** | **60 available files** |
-
-In Verified-FF, methods that accept an RGB first frame receive the same checked appearance reference for a view. In Unverified-FF, appearance is determined by the method from the prescribed text and control signals. Each multi-view case provides four synchronized third-person follow cameras. The Gunfight cases are a separate domain-specific collection used in the paper's PWM comparison; its 20 cases derive from ten source recordings.
-
-## Evaluation
-
-The paper evaluates five complementary capabilities: **controllability**, **visual and temporal quality**, **logic and state alignment**, **multi-view consistency**, and **long-horizon memory**. Its metrics include entity BBox IoU and center error; **Interaction Success Rate (ISR)**, which requires both an action and its expected end state; **Logic-Render Alignment (LRA)**, which checks actions within their assigned timeline segments; and entity reappearance and state-persistence checks for long clips. See the [paper](https://arxiv.org/abs/2610.02205) for the full protocols and model results.
-
-This repository supplies the episode inputs and records. It does **not** include generated model outputs, model weights, or the evaluator package, so the paper's scores cannot be reproduced from this repository alone.
-
-## Usage
-
-The Dataset Viewer `full` split has **600 rows, one per representation sample**. Its five group splits select rows from `full`; they are **not additional samples**. Rows contain the episode and camera IDs, representation, prompt text, and direct paths to the proxy video, prompt, first frame (when available), and manifest. The `preview_image` column contains **one image per case** (170 in total), attached to its selected `mixed` row. For 60 cases it is an available RGB first frame; for the other 110 it is an existing proxy preview. A proxy preview is **not** an RGB conditioning first frame. Browse the complete source tree in **Files and versions**.
-
-Authenticate with an account that has access to this private dataset and download it:
+## Download
 
 ```bash
+pip install -U huggingface_hub
 hf auth login
 hf download AlayaLab/PROWBench --repo-type dataset --local-dir PROWBench
 ```
 
-To fetch one proxy video:
+## Data
 
-```bash
-hf download AlayaLab/PROWBench episodes/verified_first_frame/city-push-2/views/follow/colored-obb/video.mp4 --repo-type dataset
-```
+| Split | Episodes | Duration | RGB first frames |
+|---|---:|---:|---:|
+| `verified_first_frame` | 40 | 5 s | 40 |
+| `unverified_first_frame` | 90 | 5 s | 0 |
+| `long_video` | 10 | 30 s | 0 |
+| `multi_view` | 10 | 5 s | 0 |
+| `neon` (Gunfight) | 20 | 5 s | 20 |
 
-The repository keeps the source episode layout and stores Viewer indexes separately:
+Each view has three proxy representations: `white` (coarse 3D), `mixed` (semantic proxy), and `colored-obb` (3D boxes). Multi-view episodes have four synchronized cameras. The `neon` split uses `episodes/gunfight/`.
 
-```text
-README.md
-episodes/
-  verified_first_frame/<case>/
-  unverified_first_frame/<case>/
-  long_video/<case>/
-  multi_view/<case>/
-  gunfight/<case>/
-splits/
-  <split>.parquet                  # Six Dataset Viewer indexes
-```
-
-An episode generally starts at `manifest.json` and has `scene/`, `cameras/`, `views/`, `logs/`, `annotations/`, and `validation/` as applicable. Under a selected camera's `views/<camera>/`, the `white/`, `mixed/`, and `colored-obb/` directories hold proxy media and prompts. Where supplied, the RGB appearance reference is in `views/<camera>/prompt-context/first_frame.png`. Each multi-view camera also has `views/<camera>/visibility.json`: per-frame visibility measurements for the four participants. A case may have a named version subdirectory; follow its manifest and local metadata rather than assuming all cases have identical leaves.
-
-The indexes are navigation aids; the episode files are the actual benchmark inputs. Viewer rows refer to those files and are not additional episodes or model outputs.
+The `full` Viewer split contains **600 rows**, one per camera–representation pair; the other splits are subsets. Full prompts are available through `prompt_file`. `preview_kind` distinguishes RGB first-frame references from proxy previews.
 
-## Provenance and scope
+A typical episode:
 
-The uploaded `episodes/` tree has the **same 22,821 relative file paths** as the local `PROWBench-release/episodes/` directory: the original 22,781 files plus **40 multi-view `visibility.json` files**. The six Parquet Viewer indexes are separate under the root `splits/` directory. The visibility files were copied byte-for-byte from `PROWBench-release/eval/annotations/multi4-<id>-<case>/<camera>/visibility.json` into the corresponding episode views. Their SHA-256 hashes, camera IDs, and 150-frame arrays were checked against the source evaluation bindings; all 40 have matching camera and OBB/depth inputs under the source `inputs/multi_view/` tree. `visiblePixelCount` measures pixels of the drawn participant box after depth testing, not a person segmentation mask. The source depth maps are not included here.
+```text
+episodes/<group>/<episode>/
+├── manifest.json
+├── scene/                              # Recorded world states
+├── cameras/                            # Camera records
+└── views/<camera>/
+    ├── <representation>/video.mp4
+    ├── <representation>/prompt.txt
+    ├── prompt-context/first_frame.png   # When available
+    └── visibility.json                 # Multi-view
+```
 
-To make the source directory usable by itself, **2,926 symlinks** that originally pointed to the sibling `.shared/` store were replaced with their target file bytes at the same paths, and **2,907 metadata files containing shared references** were expanded into ordinary JSON/JSONL content. Other copied files were checked against their source hashes. This is a standalone, materialized form of the episode tree; metadata files that contained shared references are not byte-identical to the compact source files.
+Long-video prompts use JSON files or a `prompts/` directory of windowed prompts. Use `prompt_file` and the adjacent `sample.json` for exact input paths, including versioned episodes.
 
-The full source bundle's root indexes, shared store, evaluation code, and historical provenance directory are outside this episodes-only release. Depth-conditioned inputs and generated model result videos are also absent. Engine event records describe the authored execution; they do not by themselves establish independent physical contact or causality. No public dataset license has been assigned to this private release.
+This release contains conditioning inputs and annotations; raw depth maps and generated model outputs are not included. [Release metadata](metadata/release.json) · [Sample metadata](metadata/samples.jsonl.gz).
 
 ## Citation
 

metadata/release.json

@@ -0,0 +1,28 @@
+{
+  "schema": "prowbench.release-metadata.v1",
+  "source_revision": "9d731e1554949a1875d79ca86023bc26a796695b",
+  "episode_file_count": 22821,
+  "episodes": 170,
+  "camera_views": 200,
+  "proxy_videos": 600,
+  "viewer_splits": {
+    "full": 600,
+    "verified_first_frame": 120,
+    "unverified_first_frame": 270,
+    "long_video": 30,
+    "multi_view": 120,
+    "neon": 60
+  },
+  "sample_metadata": "metadata/samples.jsonl.gz",
+  "sample_metadata_notes": {
+    "description": "Existing scene-context title, or manifest name when a scene-context title is unavailable.",
+    "prompt_file": "Complete prompt text is stored at this path; windowed long-video inputs are described by the adjacent sample.json.",
+    "video_matches_source": "Preserved from the original sample's prompt-binding metadata; this is not a local-versus-Hub synchronization check.",
+    "preview_kind": "rgb-first-frame denotes an available RGB reference; proxy-preview denotes a rendered conditioning preview."
+  },
+  "source_release_notes": [
+    "The uploaded `episodes/` tree has the **same 22,821 relative file paths** as the local `PROWBench-release/episodes/` directory: the original 22,781 files plus **40 multi-view `visibility.json` files**. The six Parquet Viewer indexes are separate under the root `splits/` directory. The visibility files were copied byte-for-byte from `PROWBench-release/eval/annotations/multi4-<id>-<case>/<camera>/visibility.json` into the corresponding episode views. Their SHA-256 hashes, camera IDs, and 150-frame arrays were checked against the source evaluation bindings; all 40 have matching camera and OBB/depth inputs under the source `inputs/multi_view/` tree. `visiblePixelCount` measures pixels of the drawn participant box after depth testing, not a person segmentation mask. The source depth maps are not included here.",
+    "To make the source directory usable by itself, **2,926 symlinks** that originally pointed to the sibling `.shared/` store were replaced with their target file bytes at the same paths, and **2,907 metadata files containing shared references** were expanded into ordinary JSON/JSONL content. Other copied files were checked against their source hashes. This is a standalone, materialized form of the episode tree; metadata files that contained shared references are not byte-identical to the compact source files.",
+    "The full source bundle's root indexes, shared store, evaluation code, and historical provenance directory are outside this episodes-only release. Depth-conditioned inputs and generated model result videos are also absent. Engine event records describe the authored execution; they do not by themselves establish independent physical contact or causality. No public dataset license has been assigned to this private release."
+  ]
+}

metadata/samples.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cdbb1c66c2258cf5243663b6a1006a2effd145528a44861a62bf36d63d6cfbc
+size 49431

splits/full.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c9faa5b1e109318fe7c10aabc7a4e1fd946b38a55ca47c57bec04719b6da9cf8
-size 48922896
+oid sha256:31db569b1a2bf57af37ffe8362df713fb4ddbeabfd7d2bfbffe1920fd68ab16b
+size 48231510

splits/long_video.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:db0464572fdc291141815931bca45ea2856c250b544e52085abd6c97246922a3
-size 322198
+oid sha256:e3690f4a569f7261acf66a1349d82553dcb2fc330964a6e0cb3354574abb5988
+size 286943

splits/multi_view.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9201e8ef564877a88da95f0285d27d046dff254a92274f10fd4d376e9719c6e3
-size 491158
+oid sha256:31a16e0d0bbe46be21a275910c0d6f20998e6da6dc77eb958e3d15b348b3f8f8
+size 385570

splits/neon.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:507b85ce68dec8d5be67e51496d490256e14dd785c06b52e5e5e29ff11034841
-size 9099376
+oid sha256:cf820de2b002e8c0fccd919e0cf612d45d3815a3d6bb3c8c889831cf6ae3d9e5
+size 9040459

splits/unverified_first_frame.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a83b52419ed39f8d7fef7f0e6b1df85c4c46e3da772060a19a51a98319c2da86
-size 3746287
+oid sha256:aa64e0c1245b9ee400ca3da63398cdad767a99fd82e7e92cb705555a411fd42e
+size 3416434

splits/verified_first_frame.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c10a8dafeb5b22766c529e9d1d95be6729c125b1651110fa48fec452bfe199cb
-size 35270418
+oid sha256:4db435827151359e3dfd587a1b7a1364fb26983876684d0a27f6105335e2f6d7
+size 35110966