Datasets:

PhysInOneP02
/

dataset_card_test

@@ -39,35 +39,65 @@ homepage: https://vlar-group.github.io/PhysInOne.html
 # PhysInOne: Visual Physics Learning and Reasoning in One Suite
-[Project Page](https://vlar-group.github.io/PhysInOne.html) | [Paper](https://arxiv.org/abs/2604.09415) | [Code](TODO) | [Dataset](TODO)
-## Dataset Summary
-**PhysInOne** is a large-scale synthetic dataset for visual physics learning and reasoning. It contains **153,810 dynamic 3D scenes** and **2 million annotated videos**, systematically covering **71 basic physical phenomena** across four domains of everyday physics: **mechanics**, **optics**, **fluid dynamics**, and **magnetism**.
-Unlike prior visual physics datasets that are often limited in scale, physical diversity, or annotation richness, PhysInOne provides multi-object and multi-physics interactions in complex 3D environments, together with comprehensive annotations including **geometry**, **semantics**, **motion**, **physical properties**, and **text descriptions**.
-PhysInOne is designed to support research on physics-grounded world models, physically plausible video generation, future frame prediction, physical property estimation, motion transfer, and embodied AI.
-## News
-- **2026-04-10**: PhysInOne paper released on arXiv.
-- **TODO**: Dataset released on Hugging Face.
-- **TODO**: Benchmark code and evaluation scripts released.
-## Visual Overview
 <p align="center">
   <img src="./assets/teaser.jpg" width="900">
 </p>
 <p align="center">
-  <em>PhysInOne covers diverse dynamic 3D scenes across mechanics, optics, fluid dynamics, and magnetism.</em>
 </p>
-## Dataset Examples
-The following examples show representative dynamic scenes from the four major physical domains covered by PhysInOne.
 | Mechanics                            | Fluid Dynamics                   |
 | ------------------------------------ | -------------------------------- |
@@ -77,45 +107,14 @@ The following examples show representative dynamic scenes from the four major ph
 | --------------------------------- | ------------------------------------ |
 | ![](./assets/examples_optics.gif) | ![](./assets/examples_magnetism.gif) |
-High-quality MP4 examples are also available:
-| Domain         | MP4                                           |
 | -------------- | --------------------------------------------- |
 | Mechanics      | [View video](./assets/examples_mechanics.mp4) |
 | Fluid Dynamics | [View video](./assets/examples_fluid.mp4)     |
 | Optics         | [View video](./assets/examples_optics.mp4)    |
 | Magnetism      | [View video](./assets/examples_magnetism.mp4) |
-## Dataset Highlights
-| Feature                 | Description                                                                                             |
-| ----------------------- | ------------------------------------------------------------------------------------------------------- |
-| Dynamic 3D scenes       | 153,810                                                                                                 |
-| Dynamic videos          | 2 million                                                                                               |
-| Physical phenomena      | 71                                                                                                      |
-| Physical domains        | Mechanics, Optics, Fluid Dynamics, Magnetism                                                            |
-| Multiphysics activities | 3,284                                                                                                   |
-| 3D objects              | 2,231                                                                                                   |
-| Materials               | 623                                                                                                     |
-| 3D backgrounds          | 528                                                                                                     |
-| Cameras per scene       | 12 fixed cameras + 1 moving monocular camera                                                            |
-| Video resolution        | 1120 × 1120                                                                                             |
-| Frame rate              | 30 FPS; 60 FPS for laser-related scenes                                                                 |
-| Average duration        | Approximately 5.2 seconds                                                                               |
-| Annotation types        | RGB, depth, masks, 3D trajectories, object meshes, material properties, camera poses, text descriptions |
-## Dataset Description
-PhysInOne aims to address the scarcity of large-scale, physically grounded visual data for training and evaluating AI systems. The dataset focuses on visually observable everyday physics and covers four major domains:
-- **Mechanics**: gravity, collision, acceleration, equilibrium, rotation, Hooke's law, conservation of momentum, and related phenomena.
-- **Optics**: reflection, refraction, mirror interactions, laser-related activities, and other visible optical phenomena.
-- **Fluid Dynamics**: liquid motion, splashing, buoyancy, droplets, and related interactions.
-- **Magnetism**: magnetic attraction, magnetic imbalance, and multi-object magnetic interactions.
-Each scene may involve one or more physical phenomena occurring simultaneously or sequentially. These phenomena are instantiated as concrete 3D scenes with diverse objects, materials, and backgrounds.
-## Dataset Structure
 The exact released structure may vary depending on the hosted version. A recommended structure is shown below.
@@ -124,7 +123,6 @@ PhysInOne/
 ├── README.md
 ├── assets/
 │   ├── teaser.jpg
-│   ├── dataset_statistics.jpg
 │   ├── annotation_overview.jpg
 │   ├── examples_mechanics.gif
 │   ├── examples_fluid.gif
@@ -138,39 +136,40 @@ PhysInOne/
 │   ├── train.jsonl
 │   ├── val.jsonl
 │   ├── test.jsonl
 │   ├── phenomena_taxonomy.json
-│   ├── material_properties.json
-│   └── benchmark_subsets.json
 ├── videos/
 │   ├── train/
 │   ├── val/
 │   └── test/
 ├── annotations/
 │   ├── depth/
-│   ├── masks/
 │   ├── trajectories/
-│   ├── camera_poses/
-│   ├── meshes/
-│   └── material_properties/
 └── scripts/
     ├── download.py
     ├── load_sample.py
     └── visualize_sample.py
 ```
-Each scene is associated with multiple rendered videos and annotations. A typical metadata entry may look like:
 ```json
 {
-  "scene_id": "scene_000000",
-  "activity_id": "activity_0000",
   "split": "train",
   "physical_domains": ["mechanics"],
   "physical_phenomena": ["gravity", "collision"],
-  "caption": "A red ball rolls down a sloped surface and collides with a box...",
-  "duration": 5.2,
-  "fps": 30,
-  "resolution": [1120, 1120],
   "videos": {
     "fixed_camera_00": "videos/train/scene_000000/camera_00.mp4",
     "fixed_camera_01": "videos/train/scene_000000/camera_01.mp4",
@@ -178,313 +177,439 @@ Each scene is associated with multiple rendered videos and annotations. A typica
   },
   "annotations": {
     "depth": "annotations/depth/train/scene_000000/",
-    "masks": "annotations/masks/train/scene_000000/",
-    "trajectories": "annotations/trajectories/train/scene_000000.json",
-    "camera_poses": "annotations/camera_poses/train/scene_000000.json",
-    "meshes": "annotations/meshes/train/scene_000000/",
-    "material_properties": "annotations/material_properties/train/scene_000000.json"
   }
 }
 ```
-## Data Splits
-PhysInOne is split into training, validation, and held-out test sets with an **8:1:1** ratio. To reduce data leakage, 3D assets are assigned exclusively to one split.
-| Split         | Ratio | Description                                                                 |
-| ------------- | -----:| --------------------------------------------------------------------------- |
-| Train         | 80%   | Used for model training and fine-tuning                                     |
-| Validation    | 10%   | Used for validation, ablation, and model selection                          |
-| Held-out Test | 10%   | Used for final evaluation; 3D assets are disjoint from train and validation |
-## Benchmark Subsets
-The paper uses several smaller subsets for efficient benchmarking.
-| Subset                 | Size                  | Used for                                         |
-| ---------------------- | ---------------------:| ------------------------------------------------ |
-| `test-small`           | 772 text-video pairs  | Physics-aware video generation                   |
-| `test-mini`            | 103 scenes            | Long-term and short-term future frame prediction |
-| `test-tiny`            | 20 scenes             | Physical property estimation                     |
-| Motion transfer subset | 273 validation scenes | Motion transfer evaluation                       |
-## Data Fields
-The dataset may contain the following fields.
-| Field                 | Type         | Description                                                              |
-| --------------------- | ------------ | ------------------------------------------------------------------------ |
-| `scene_id`            | string       | Unique identifier of a dynamic 3D scene                                  |
-| `activity_id`         | string       | Identifier of the physical activity                                      |
-| `split`               | string       | Dataset split: train, validation, or test                                |
-| `physical_domains`    | list[string] | Physics domains involved in the scene                                    |
-| `physical_phenomena`  | list[string] | Basic physical phenomena involved in the scene                           |
-| `caption`             | string       | English description of the scene, visual elements, and physical activity |
-| `videos`              | dict         | Paths to videos rendered from fixed and moving cameras                   |
-| `depth`               | path         | Ground-truth depth images                                                |
-| `masks`               | path         | Per-frame object masks                                                   |
-| `trajectories`        | path/json    | 3D trajectories of dynamic objects                                       |
-| `camera_poses`        | path/json    | Camera intrinsics and extrinsics                                         |
-| `meshes`              | path         | Object mesh files                                                        |
-| `material_properties` | path/json    | Physical and material properties of scene objects                        |
-## Annotation Details
-PhysInOne provides synchronized visual and physical annotations for each dynamic 3D scene.
-<p align="center">
-  <img src="./assets/annotation_overview.jpg" width="900">
-</p>
-| Annotation          | Description                                                                                                      |
-| ------------------- | ---------------------------------------------------------------------------------------------------------------- |
-| RGB videos          | Rendered videos from 12 fixed cameras and 1 moving monocular camera                                              |
-| Depth maps          | Ground-truth depth images synchronized with rendered RGB frames                                                  |
-| Object masks        | Per-frame object masks for semantic or instance-level scene understanding                                        |
-| 3D trajectories     | Object-level motion trajectories over time                                                                       |
-| Camera poses        | Camera parameters for multi-view geometric reasoning                                                             |
-| Object meshes       | 3D mesh assets associated with scene objects                                                                     |
-| Material properties | Physical and material parameters such as density, friction, restitution, and other simulation-related properties |
-| Text descriptions   | Human-written and proofread English paragraphs describing the scene and physical activity                        |
-Users should check the released metadata specification for coordinate systems, units, camera conventions, mask encoding, and material parameter definitions.
-## Supported Tasks and Benchmarks
-PhysInOne is designed to support several visual physics learning and reasoning tasks.
-### 1. Physics-aware Video Generation
-Given text prompts, initial frames, or image conditions, models are expected to generate videos that are not only visually realistic but also physically plausible.
-Potential settings include:
-- Text-to-video generation
-- Image-to-video generation
-- Text-image-to-video generation
-- Video model fine-tuning with physics-rich data
-Representative models evaluated in the paper include:
-- SVD-XT
-- CogVideoX-1.5-5B
-- Wan2.2-5B
-Suggested evaluation metrics include:
-- PMF: Physical Motion Fidelity
-- FVD
-- Human physical plausibility rating
-### 2. Long-term Future Frame Prediction
-Given the first half of a dynamic scene, models are expected to predict the second half of the video while preserving physically plausible motion.
-This setting can be used to evaluate:
-- Scene-specific 4D modeling methods
-- Video prediction models
-- Seen-view prediction
-- Novel-view prediction
-Representative models evaluated in the paper include:
-- TiNeuVox
-- DefGS
-- TRACE
-- FreeGave
-- ExtDM
-- MAGI-1
-Suggested evaluation metrics include:
-- PMF
-- PSNR
-- SSIM
-- LPIPS
-### 3. Continuous Short-term Future Frame Prediction
-Given a stream of observed frames, models are expected to continuously predict the next few frames in a real-time or near-real-time manner.
-This setting is useful for:
-- Future-aware robot planning
-- Embodied AI
-- Dynamic scene understanding
-- Short-horizon physical prediction
-Representative models evaluated in the paper include:
-- DefGS
-- FreeGave
-- ExtDM
-- MAGI-1
-Suggested evaluation metrics include:
-- PMF
-- PSNR
-- SSIM
-- LPIPS
-### 4. Physical Property Estimation
-Given visual observations of dynamic scenes, models are expected to infer physical properties of objects and materials.
-Target properties may include:
-- Young's modulus
-- Poisson's ratio
-- Viscosity
-- Bulk modulus
-- Yield stress
-- Friction angle
-- Initial velocity
-Representative models evaluated in the paper include:
-- PAC-NeRF
-- GIC
-This task is useful for inverse physics, system identification, resimulation, editable scene dynamics, and robot interaction.
-### 5. Motion Transfer
-Given a source video and a target image or scene, models are expected to transfer physically meaningful motion patterns from the source to the target while preserving the target appearance.
-Representative models evaluated in the paper include:
-- MotionPro
-- GoWithTheFlow
-Suggested evaluation metrics include:
-- PMF
-- PSNR
-- SSIM
-- LPIPS
-## How to Use
-### Install Dependencies
-```bash
-pip install datasets huggingface_hub
-```
-### Download Metadata Only
-```bash
-huggingface-cli download TODO/PhysInOne \
-  --include "metadata/*" \
-  --local-dir ./PhysInOne
 ```
-### Download a Small Subset
-```bash
-huggingface-cli download TODO/PhysInOne \
-  --include "metadata/*" \
-  --include "videos/test-small/*" \
-  --local-dir ./PhysInOne
 ```
-### Load with `datasets`
-```python
-from datasets import load_dataset
-dataset = load_dataset("TODO/PhysInOne", split="train")
-sample = dataset[0]
-print(sample.keys())
-print(sample["caption"])
-print(sample["physical_domains"])
-print(sample["physical_phenomena"])
 ```
-### Visualize a Sample
-```python
-import json
-from pathlib import Path
-metadata_path = Path("./PhysInOne/metadata/train.jsonl")
-with open(metadata_path, "r") as f:
-    sample = json.loads(next(f))
-print("Scene ID:", sample["scene_id"])
-print("Caption:", sample["caption"])
-print("Videos:", sample["videos"])
-print("Annotations:", sample["annotations"])
 ```
-## Data Creation Pipeline
-PhysInOne is constructed through a multi-stage pipeline.
-### 1. Identifying Physical Phenomena and Laws
-The dataset identifies 71 basic physical phenomena across mechanics, optics, fluid dynamics, and magnetism. These phenomena are governed by fundamental physical laws such as Newton's laws, conservation of momentum, Hooke's law, Bernoulli's principle, and related laws.
-### 2. Collecting 3D Assets
-The dataset uses diverse 3D objects, materials, and backgrounds to instantiate realistic physical scenes. These assets include solid objects, interactable objects, destructible objects, deformable objects, granular objects, liquids, materials, and 3D backgrounds.
-### 3. Creating Multiphysics Multi-object 3D Scenes
-Basic physical phenomena are combined into single-, double-, and triple-physics activities. These conceptual activities are then instantiated as concrete 3D scenes with diverse objects, materials, and backgrounds.
-### 4. Simulating Physical Dynamics
-Multiple simulation tools and engines are used to simulate different types of physical dynamics. Chaos Physics in Unreal Engine is used for many everyday rigid-body physical phenomena. MPM is used for deformable and granular objects, and SPH is used for liquid simulations.
-### 5. Rendering and Annotation
-Each scene is rendered from 12 fixed cameras and one moving monocular camera. During rendering, the dataset generates RGB videos and ground-truth annotations including depth, masks, 3D trajectories, camera poses, meshes, material properties, and text descriptions.
-## Intended Uses
-PhysInOne is intended for research on:
-- Physics-aware video generation
-- World models and physical reasoning
-- Future frame prediction
-- Multi-view dynamic scene understanding
-- Physical property estimation
-- Inverse physics and system identification
-- Motion transfer
-- Embodied AI and robot learning
-- Simulation-aware visual representation learning
-## Out-of-Scope Uses
-PhysInOne is not intended to be used as:
-- A perfect replacement for real-world physical measurements
-- A safety-critical physics simulator
-- A dataset for non-visual physics such as thermodynamics or acoustics
-- A benchmark for domains not covered by the dataset
-- A source of guaranteed real-world physical fidelity without validation
-## Limitations
-PhysInOne is a synthetic dataset. Although the scenes are generated with physics simulation and controlled rendering, users should be aware of several limitations:
-- **Synthetic-to-real gap**: Models trained on PhysInOne may require adaptation before deployment on real-world videos.
-- **Simulator approximation**: Current simulators may not perfectly reproduce all real-world physical effects.
-- **Asset distribution bias**: The dataset distribution depends on the collected 3D objects, materials, and backgrounds.
-- **Physics coverage**: PhysInOne focuses on visually observable everyday physics in mechanics, optics, fluid dynamics, and magnetism. It does not cover all areas of physics.
-- **Rendering bias**: The visual appearance is influenced by the rendering engine, lighting, materials, and asset sources.
-- **Storage cost**: The full dataset contains 2 million videos and may require substantial storage and bandwidth.
 ## License
 TODO: Add the final dataset license.
-Please also check the licenses of third-party 3D assets, materials, and backgrounds used in the dataset. If individual assets have separate licenses, they should be documented clearly.
 ## Citation
@@ -505,6 +630,7 @@ For questions about the dataset, please contact:
 - TODO: contact email
 - Project page: https://vlar-group.github.io/PhysInOne.html
 ## Acknowledgements

 # PhysInOne: Visual Physics Learning and Reasoning in One Suite
+<p align="center">
+  <a href="https://vlar-group.github.io/PhysInOne.html">Project Page</a> |
+  <a href="https://arxiv.org/abs/2604.09415">Paper</a> |
+  <a href="TODO">Code</a> |
+  <a href="TODO">Dataset</a> |
+  <a href="TODO">Leaderboard</a>
+</p>
+## Dataset Card
+### Title, Authors, and Conference Information
+**Title:** PhysInOne: Visual Physics Learning and Reasoning in One Suite
+**Authors:** Siyuan Zhou, Hejun Wang, Hu Cheng, Jinxi Li, Dongsheng Wang, Junwei Jiang, Yixiao Jin, Jiayue Huang, Shiwei Mao, Shangjia Liu, Yafei Yang, Hongkang Song, Shenxing Wei, Zihui Zhang, Bing Wang, Zhihua Wang, Chuhang Zou, Bo Yang, and DataTeam.
+**Affiliations:** vLAR Group, The Hong Kong Polytechnic University, Syai Singapore, Meta.
+**Conference / Venue:** TODO.
+**Paper:** https://arxiv.org/abs/2604.09415
+**Project Page:** https://vlar-group.github.io/PhysInOne.html
+### Summary
+**PhysInOne** is a large-scale synthetic dataset for visual physics learning and reasoning. It contains **153,810 dynamic 3D scenes** and **2 million annotated videos**, covering **71 basic physical phenomena** across four domains of everyday physics: **mechanics**, **optics**, **fluid dynamics**, and **magnetism**.
+Each scene may contain multi-object and multi-physics interactions in complex 3D environments. PhysInOne provides rich annotations including RGB videos, depth maps, object masks, 3D trajectories, camera poses, object meshes, material properties, and textual descriptions.
+The dataset is designed to support research on physics-aware video generation, future frame prediction, physical property estimation, motion transfer, physical reasoning, and world models.
+### News and Release Timetable
+| Component                       | Status                    | Expected Release |
+| ------------------------------- | ------------------------- | ---------------- |
+| Rendered data                   | Partially released / TODO | TODO             |
+| Rendered data: train split      | TODO                      | TODO             |
+| Rendered data: test split       | TODO                      | TODO             |
+| Rendered data: validation split | TODO                      | TODO             |
+| Train split update              | TODO                      | May 21           |
+| 3D assets                       | Not yet released          | Around June      |
+| Leaderboard                     | Ongoing                   | Link TODO        |
+| Baseline code                   | Not yet released          | Around June      |
+| Data processing code            | Not yet released          | Around June      |
+### Visual Overview
 <p align="center">
   <img src="./assets/teaser.jpg" width="900">
 </p>
 <p align="center">
+  <em>PhysInOne covers dynamic 3D physical scenes across mechanics, fluid dynamics, optics, and magnetism.</em>
 </p>
+### Video Demo Gallery
+We provide lightweight GIF previews in this README and high-quality MP4 examples as separate files.
 | Mechanics                            | Fluid Dynamics                   |
 | ------------------------------------ | -------------------------------- |
 | --------------------------------- | ------------------------------------ |
 | ![](./assets/examples_optics.gif) | ![](./assets/examples_magnetism.gif) |
+| Domain         | High-quality MP4                              |
 | -------------- | --------------------------------------------- |
 | Mechanics      | [View video](./assets/examples_mechanics.mp4) |
 | Fluid Dynamics | [View video](./assets/examples_fluid.mp4)     |
 | Optics         | [View video](./assets/examples_optics.mp4)    |
 | Magnetism      | [View video](./assets/examples_magnetism.mp4) |
+### Dataset Structure
 The exact released structure may vary depending on the hosted version. A recommended structure is shown below.
 ├── README.md
 ├── assets/
 │   ├── teaser.jpg
 │   ├── annotation_overview.jpg
 │   ├── examples_mechanics.gif
 │   ├── examples_fluid.gif
 │   ├── train.jsonl
 │   ├── val.jsonl
 │   ├── test.jsonl
+│   ├── benchmark_subsets.json
 │   ├── phenomena_taxonomy.json
+│   └── material_properties.json
 ├── videos/
 │   ├── train/
 │   ├── val/
 │   └── test/
 ├── annotations/
 │   ├── depth/
+│   ├── segmentation/
+│   ├── captions/
 │   ├── trajectories/
+│   ├── cameras/
+│   └── pointclouds/
 └── scripts/
     ├── download.py
     ├── load_sample.py
     └── visualize_sample.py
 ```
+A typical metadata item may look like:
 ```json
 {
+  "id": "scene_000000",
+  "scene_name": "TODO",
   "split": "train",
+  "activity_type": "single",
   "physical_domains": ["mechanics"],
   "physical_phenomena": ["gravity", "collision"],
+  "caption_path": "annotations/captions/train/scene_000000/caption.txt",
+  "ue_path": "TODO/scene_000000",
+  "repo_link": "TODO",
+  "download_link": "TODO",
   "videos": {
     "fixed_camera_00": "videos/train/scene_000000/camera_00.mp4",
     "fixed_camera_01": "videos/train/scene_000000/camera_01.mp4",
   },
   "annotations": {
     "depth": "annotations/depth/train/scene_000000/",
+    "segmentation": "annotations/segmentation/train/scene_000000/",
+    "caption": "annotations/captions/train/scene_000000/caption.txt",
+    "trajectory": "annotations/trajectories/train/scene_000000/trajectory.json",
+    "camera": "annotations/cameras/train/scene_000000/camera.json",
+    "pointcloud": "annotations/pointclouds/train/scene_000000/points.ply"
   }
 }
 ```
+### Data Splits
+PhysInOne is split into train, validation, and test sets. Each split is intended for a different stage of model development and evaluation.
+| Split      | Purpose                      | Description                                                    |
+| ---------- | ---------------------------- | -------------------------------------------------------------- |
+| Train      | Training and fine-tuning     | Used for learning from rendered dynamic scenes and annotations |
+| Validation | Model selection and ablation | Used for validation and development-time evaluation            |
+| Test       | Final evaluation             | Used for held-out benchmarking and leaderboard submission      |
+The split metadata files are expected to be provided as:
+```text
+metadata/train.jsonl
+metadata/val.jsonl
+metadata/test.jsonl
+```
+Each JSONL entry should contain the scene identifier, scene name, physical category labels, file paths, and download links.
+## Dataset Viewer
+The Hugging Face Dataset Viewer is designed to help users quickly search, filter, and export scene-level metadata without downloading the full dataset.
+### Filtering and Search
+The viewer should support filtering by:
+- **Activity complexity:** `single`, `double`, `triple`
+- **Physical domain:** `mechanics`, `fluid_dynamics`, `optics`, `magnetism`
+- **Keyword search:** scene name, physical phenomenon, caption keyword, object keyword, or material keyword
+- **Split:** `train`, `val`, `test`
+- **Availability:** rendered data, 3D assets, annotations, benchmark subset
+### Viewer Table
+After selecting filters, the viewer should display a dataframe-like table with at least the following columns:
+| Column            | Description                                                   |
+| ----------------- | ------------------------------------------------------------- |
+| `id`              | Unique scene identifier                                       |
+| `scene_name`      | Human-readable scene name                                     |
+| `split`           | Train / validation / test split                               |
+| `activity_type`   | Single-, double-, or triple-physics activity                  |
+| `physical_domain` | Mechanics, fluid dynamics, optics, or magnetism               |
+| `phenomena`       | Physical phenomena involved in the scene                      |
+| `ue_path`         | Unreal Engine scene or asset path                             |
+| `repo_link`       | Link to the corresponding repository item or hosted data page |
+| `download_link`   | Direct download link for the scene package or rendered data   |
+### JSON Export
+The viewer should provide an **Export JSON** button. The exported JSON should contain selected scenes and their download links.
+Example export format:
+```json
+{
+  "selected_scenes": [
+    {
+      "id": "scene_000000",
+      "scene_name": "TODO",
+      "download_link": "TODO"
+    },
+    {
+      "id": "scene_000001",
+      "scene_name": "TODO",
+      "download_link": "TODO"
+    }
+  ]
+}
+```
+This JSON file can be passed directly to the download script.
+## How to Use
+### Install Dependencies
+```bash
+pip install datasets huggingface_hub pandas tqdm
+```
+### Download Metadata Only
+```bash
+huggingface-cli download TODO/PhysInOne \
+  --include "metadata/*" \
+  --local-dir ./PhysInOne
+```
+### Download by Exported JSON
+After selecting scenes in the Dataset Viewer, export the selected scene list as JSON and download the corresponding files:
+```bash
+python scripts/download.py \
+  --selection selected_scenes.json \
+  --output_dir ./PhysInOne
+```
+### Download a Split
+```bash
+python scripts/download.py \
+  --split train \
+  --output_dir ./PhysInOne
+```
+```bash
+python scripts/download.py \
+  --split val \
+  --output_dir ./PhysInOne
+```
+```bash
+python scripts/download.py \
+  --split test \
+  --output_dir ./PhysInOne
+```
+### Load Metadata
+```python
+import json
+from pathlib import Path
+metadata_path = Path("./PhysInOne/metadata/train.jsonl")
+with open(metadata_path, "r") as f:
+    sample = json.loads(next(f))
+print("ID:", sample["id"])
+print("Scene name:", sample["scene_name"])
+print("Split:", sample["split"])
+print("Physical domains:", sample["physical_domains"])
+print("Phenomena:", sample["physical_phenomena"])
+print("Download link:", sample["download_link"])
+```
+### Visualize a Scene
+```bash
+python scripts/visualize_sample.py \
+  --scene_id scene_000000 \
+  --data_root ./PhysInOne
+```
+## Benchmark Subsets
+We provide mini benchmark subsets for lightweight evaluation and quick prototyping.
+| Subset                | Size                  | Intended Use                                     |
+| --------------------- | ---------------------:| ------------------------------------------------ |
+| `test-small`          | 772 text-video pairs  | Physics-aware video generation                   |
+| `test-mini`           | 103 scenes            | Long-term and short-term future frame prediction |
+| `test-tiny`           | 20 scenes             | Physical property estimation                     |
+| `motion-transfer-val` | 273 validation scenes | Motion transfer evaluation                       |
+The benchmark subset metadata is expected to be stored in:
+```text
+metadata/benchmark_subsets.json
+```
+Example format:
+```json
+{
+  "test-small": ["scene_000000", "scene_000001"],
+  "test-mini": ["scene_000100", "scene_000101"],
+  "test-tiny": ["scene_000200", "scene_000201"],
+  "motion-transfer-val": ["scene_000300", "scene_000301"]
+}
+```
+## Data Fields
+### Abbreviations
+| Abbreviation | Meaning                                     |
+| ------------ | ------------------------------------------- |
+| `id`         | Unique scene identifier                     |
+| `ue_path`    | Unreal Engine project, scene, or asset path |
+| `repo_link`  | Link to the corresponding hosted repo item  |
+| `rgb`        | Rendered RGB video                          |
+| `depth`      | Ground-truth depth                          |
+| `seg`        | Segmentation mask                           |
+| `traj`       | Object trajectory                           |
+| `cam`        | Camera metadata                             |
+| `pc`         | Point cloud                                 |
+### Core Fields
+| Field                | Type         | Description                                                                      |
+| -------------------- | ------------ | -------------------------------------------------------------------------------- |
+| `id`                 | string       | Unique scene identifier                                                          |
+| `scene_name`         | string       | Human-readable scene name                                                        |
+| `split`              | string       | `train`, `val`, or `test`                                                        |
+| `activity_type`      | string       | `single`, `double`, or `triple`                                                  |
+| `physical_domains`   | list[string] | One or more of mechanics, fluid dynamics, optics, and magnetism                  |
+| `physical_phenomena` | list[string] | Physical phenomena involved in the scene                                         |
+| `caption_path`       | string       | Path to `caption.txt`                                                            |
+| `ue_path`            | string       | Unreal Engine scene or asset path                                                |
+| `repo_link`          | string       | Hosted repository link                                                           |
+| `download_link`      | string       | Scene-level or package-level download link                                       |
+| `videos`             | dict         | Paths to fixed-camera and moving-camera videos                                   |
+| `annotations`        | dict         | Paths to depth, segmentation, caption, trajectory, camera, and point cloud files |
+## Annotation Details
+PhysInOne provides synchronized visual and physical annotations for each dynamic 3D scene.
+<p align="center">
+  <img src="./assets/annotation_overview.jpg" width="900">
+</p>
+### Depth
+Depth maps are synchronized with RGB frames.
+Please specify the following in the final release:
+- Depth unit: TODO, for example meter or Unreal Engine unit.
+- Depth convention: TODO, for example camera-space z-depth or Euclidean distance.
+- File format: TODO, for example PNG, EXR, NPY, or NPZ.
+- Value range and invalid value convention: TODO.
+### Segmentation
+Segmentation masks encode background, static foreground objects, and dynamic foreground objects.
+Expected encoding:
+| Pixel Value | Meaning                    |
+| -----------:| -------------------------- |
+| `0`         | Background                 |
+| `1-127`     | Static foreground objects  |
+| `128-255`   | Dynamic foreground objects |
+Please specify whether the segmentation is semantic-level, instance-level, or mixed.
+### Captions
+Each scene includes a `caption.txt` file containing an English paragraph that describes the visual elements and the physical activity.
+Expected path format:
+```text
+annotations/captions/{split}/{scene_id}/caption.txt
 ```
+### Trajectories
+Each scene includes a `trajectory.json` file storing object trajectory data.
+Recommended contents:
+```json
+{
+  "scene_id": "scene_000000",
+  "fps": 30,
+  "objects": [
+    {
+      "object_id": "object_000",
+      "object_name": "TODO",
+      "is_dynamic": true,
+      "positions": [[0.0, 0.0, 0.0]],
+      "rotations": [[1.0, 0.0, 0.0, 0.0]],
+      "timestamps": [0.0]
+    }
+  ]
+}
 ```
+Please specify the coordinate system, units, rotation convention, and timestamp convention in the final release.
+### Cameras
+Each scene includes a camera JSON file describing fixed cameras and the moving monocular camera.
+Recommended contents:
+```json
+{
+  "scene_id": "scene_000000",
+  "cameras": [
+    {
+      "camera_id": "camera_00",
+      "type": "fixed",
+      "intrinsics": {
+        "fx": 0.0,
+        "fy": 0.0,
+        "cx": 0.0,
+        "cy": 0.0,
+        "width": 1120,
+        "height": 1120
+      },
+      "extrinsics": {
+        "world_to_camera": [[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]]
+      }
+    }
+  ]
+}
 ```
+Please specify whether camera extrinsics are stored as world-to-camera or camera-to-world matrices.
+### Point Clouds
+Each scene may include `points.ply`.
+Expected behavior:
+- The point cloud is sampled from the first frame.
+- It includes multiple camera views after depth back-projection.
+- It is randomly sampled to approximately **100,000 points**.
+- The file format is `.ply`.
+Recommended path format:
+```text
+annotations/pointclouds/{split}/{scene_id}/points.ply
 ```
+Please specify whether point colors, normals, semantic labels, or instance labels are included in the `.ply` file.
+## Supported Tasks and Benchmarks
+PhysInOne supports the following visual physics learning and reasoning tasks.
+### Physics-aware Video Generation
+Given text prompts, image conditions, or initial frames, models generate videos that should be visually realistic and physically plausible.
+Representative settings:
+- Text-to-video generation
+- Image-to-video generation
+- Text-image-to-video generation
+- Video model fine-tuning with physics-rich data
+Suggested metrics:
+- PMF: Physical Motion Fidelity
+- FVD
+- Human physical plausibility rating
+### Long-term Future Frame Prediction
+Given the first half of a dynamic scene, models predict the second half of the video.
+Representative settings:
+- Seen-view prediction
+- Novel-view prediction
+- Scene-specific 4D modeling
+- Video prediction
+Suggested metrics:
+- PMF
+- PSNR
+- SSIM
+- LPIPS
+### Continuous Short-term Future Frame Prediction
+Given streaming observations, models continuously predict the next few frames.
+This setting is useful for:
+- Future-aware robot planning
+- Embodied AI
+- Short-horizon physical prediction
+- Dynamic scene understanding
+Suggested metrics:
+- PMF
+- PSNR
+- SSIM
+- LPIPS
+### Physical Property Estimation
+Given visual observations, models estimate physical properties of scene objects and materials.
+Example target properties:
+- Young's modulus
+- Poisson's ratio
+- Viscosity
+- Bulk modulus
+- Yield stress
+- Friction angle
+- Initial velocity
+### Motion Transfer
+Given a source video and a target image or target scene, models transfer physically meaningful motion patterns while preserving the target appearance.
+Suggested metrics:
+- PMF
+- PSNR
+- SSIM
+- LPIPS
 ## License
 TODO: Add the final dataset license.
+Please verify the license terms for:
+- Rendered RGB videos
+- Annotations
+- 3D assets
+- Materials
+- Backgrounds
+- Code
+- Benchmark metadata
+If third-party assets have separate licenses, please document them clearly.
 ## Citation
 - TODO: contact email
 - Project page: https://vlar-group.github.io/PhysInOne.html
+- Hugging Face dataset page: TODO
 ## Acknowledgements