Update README.md

README.md

@@ -6,11 +6,10 @@ pretty_name: >-
 size_categories:
 - 100K<n<1M
 task_categories:
-- reinforcement-learning
-- robotics
-- world models
-- representation-learning
-- video-understanding
+  - reinforcement-learning
+  - representation-learning
+  - video-understanding
+
 tags:
   - egocentric
   - robotics
@@ -24,57 +23,72 @@ tags:
 ---
 
 
-# Fidelity Dynamics – Egocentric State–Action Transitions (v0)
+# Fidelity Data Factory – Egocentric State–Action Transitions (v0)
+
+This repository contains an initial release of structured state–action–state′
+transitions extracted from real-world egocentric video.
 
-This repository contains an initial release of egocentric state–action–state′ transitions extracted from real-world worker footage by buildai.
+The goal of this dataset is to provide early infrastructure for learning
+dynamics and representations from large-scale human activity data.
 
 ## Overview
-The dataset is constructed by enriching monocular egocentric videos into structured transitions of the form:
+
+Each data point is a short temporal transition of the form:
 
 (s_t, a_t, s_{t+1})
 
-Each transition represents a short temporal step derived from consecutive video frames.
+Transitions are derived from monocular egocentric footage recorded in real
+factory environments.
 
-This release is intended as early infrastructure for researchers exploring:
-- World models
-- Dynamics learning
-- Vision–language–action systems
-- Representation learning from human activity
+This release does not include robot-specific signals such as torques or joint
+states, and is intended for research and exploration rather than deployment.
 
 ## Data Contents
-- ~250,000 transitions
-- Real factory environments
-- Egocentric viewpoint (head- or chest-mounted cameras)
 
-Each transition is stored as a JSONL record.
+- ~200k+ transitions
+- Egocentric (head / chest-mounted) viewpoint
+- Real industrial environments
+
+Transitions are stored in JSONL format.
 
 ## Schema (Simplified)
 
-- `s` (state):
-  - `ego.pose`: estimated egomotion
-  - `ego.vel`: egocentric velocity
-  - `hand`: hand presence and image-space location
-  - `entities`: detected objects with bounding boxes and centers
-  - `meta`: video identifier
+Each record contains:
+
+- `s`:
+  - `ego_pose`
+  - `ego_velocity`
+  - `hand_state`
+  - `entities` (objects with image-space location)
+  - `meta` (video id, timestamp)
 
-- `a` (action):
+- `a`:
   - `ego_delta`
   - `hand_delta`
-  - `grasp_delta`
+  - `interaction_delta`
 
 - `s_prime`:
   - Same structure as `s`, representing the next timestep
 
-All values are derived from monocular video without force, torque, or privileged robot sensors.
+See `schema.json` for full details.
 
-## Notes & Limitations
-- Monocular only
-- No force / torque / joint states
-- No task labels
-- Noise and estimation error are expected
+## Intended Use
+
+This dataset may be useful for:
+- World model research
+- Offline RL
+- Vision–language–action pretraining
+- Learning dynamics from human activity
+- Representation learning from egocentric video
 
-This is an early dataset released to support exploration and feedback.
+## Limitations
+
+- Monocular video only
+- No force / torque signals
+- No task labels
+- Contains estimation noise
 
 ## Credits
-Original video data provided by BuildAI.
-Enrichment and processing by Fidelity Dynamics.
+
+Original video data provided by BuildAI.  
+Enrichment and processing by Fidelity Dynamics.