EgoExo-Wreck: 150+ Hours of Synchronized Egocentric/Exocentric Video — Object Destruction, Debris Manipulation & Cleanup

EgoExo-Wreck is a synchronized egocentric + exocentric (first-person + third-person) video dataset captured at a commercial rage room facility, produced specifically for AI training. It documents two rare, physics-rich activity domains: high-force human-object interaction — real participants destroying electronics, furniture, glassware, and ceramics with hand tools (bats, crowbars, hammers) — and debris cleanup and environment reset — trained staff sweeping, collecting fragments, shoveling debris, and re-placing objects across repeated procedural cycles (destroy → clean → reset).

Its significance: paired ego-exo capture with commercial AI-training rights are scarce, and the content covers manipulation of broken, irregular, and deformable objects in unstructured environments — object classes largely absent from household-activity datasets — plus real-world impact and shattering dynamics valuable for physics and world-model research. Every participant has signed explicit AI-training and third-party-licensing consent with a session-to-signature audit trail. The corpus grows ~80 hours per month through ongoing directed capture.

Data Product Features

• Paired video streams: head-mounted egocentric 4K (GoPro Hero 12, 3840×2160 @ 30/60 fps) + fixed-mount exocentric (2688×1520 or 1920×1080 @ 30 fps), per session
• Per-clip synchronization data: ego/exo sync frame anchors, lag in seconds, confidence rating, and method (audio waveform correlation, 8 kHz mono RMS-delta envelope; ±~50 ms accuracy)
• Machine-readable manifest (CSV): clip IDs, session IDs, capture dates, activity class, room/location, durations, synchronized-overlap runtime, device metadata, frame rates, file sizes
• Object-detection auto-labels: 16-class in-domain detector (room object inventory including tools), with documented training methodology and per-class metrics
• Multi-person sessions: 1–6 participants per destruction session, including multi-agent coordination footage

Distribution

• Format: MP4 video (H.264/H.265); CSV manifest; label files (YOLO/COCO format); narration transcripts (text with timestamps)
• Data Volume: 150 hours of synchronized ego-exo video across paired clips (per-clip synchronized-overlap runtime documented in manifest); growing ~80 hours/month; typical clip length 6–31 minutes; ego files ~1.5–20 GB per clip, exo files ~0.4–1 GB per clip
• Structure: session-based pairing — each record links one egocentric file to its matched exocentric file with frame-level sync anchors
• Delivery: secure file transfer or cloud bucket; evaluation sample available under click-through agreement

Usage

This data product is ideal for a variety of applications:

• Robot manipulation / imitation learning: demonstrations of grasping, sweeping, and handling irregular, deformable, and fragmented objects (broken glass, ceramic shards, mixed debris) in unstructured settings
• Vision-language-action (VLA) model training: paired ego-exo demonstrations for instruction-conditioned learning
• World models / video generation: real impact dynamics, shattering, and object state transitions (intact → damaged → fragment) from synchronized viewpoints
• Cross-view understanding: ego-exo correspondence, view-invariant action recognition, cross-view retrieval
• Activity recognition and temporal segmentation: repeated procedural cycles with consistent scenes and natural variation
• Physics and damage modeling: before/during/after sequences of object destruction for simulation ground truth and damage-assessment models

Coverage

• Geographic Coverage: United States (single facility) — 4 destruction rooms plus warehouse/storage reset areas
• Time Range: January 2026 – present (ongoing monthly capture)
• Demographics: adult participants, 1–6 per session; staff-performed cleanup sequences; demographic attributes not annotated

License

Proprietary

AI Training Rights

Licensee is granted a non-exclusive, worldwide, and perpetual right to:

• Use the Data Product to train, fine-tune, and evaluate machine learning models, including large language models.
• Incorporate Data Product content into models and commercialize resulting model outputs.
• Create derivative works (model weights, embeddings, etc.) for any lawful purpose.

Restrictions:

• The Data Product itself may not be sold, redistributed, or shared outside of licensed usage.
• Licensee must comply with all applicable laws, including data protection and privacy regulations.

Who Can Use It

• Robotics and embodied-AI labs: training manipulation policies and VLA models on human demonstrations of debris handling and object interaction
• World-model and video-generation teams: learning physical dynamics from real destruction and state-transition footage
• Computer vision researchers: ego-exo benchmarks, cross-view correspondence, action recognition, temporal segmentation
• Simulation, game, and VFX developers: real shatter and impact reference for physics simulation ground truth
• Insurance and damage-assessment AI developers: before/after damage sequences for assessment-model training

Data Dictionary

Column Name	Data Type	Description	Possible Values/Notes
Record ID	string	Sequential record identifier	001–012 (sample); zero-padded
Session ID	string	Capture session identifier	Format S{YYYY}-{MM}
Clip ID	string	Unique clip identifier	Format PM-{X|H}-YYYY-MM-NNN; X = destruction, H = cleanup
Capture Date	date	Date of capture session	ISO 8601 (YYYY-MM-DD)
Activity	string	Activity class description	"High-force object destruction" or "Debris cleanup / reset manipulation"
Activity Code	string	Machine-filterable activity class	DESTRUCT, CLEANUP
Location	string	Capture location within facility	Room 1–4, Storage/Warehouse
Ego Duration	time	Egocentric clip runtime	HH:MM:SS
Exo Duration	time	Exocentric clip runtime	HH:MM:SS
Synced Overlap	time	Runtime where both views exist (usable paired footage)	HH:MM:SS
Synced Overlap (min)	float	Same, in decimal minutes for aggregation	e.g., 27.5
Ego Device	string	Egocentric camera model	GoPro Hero 12
Ego Resolution	string	Egocentric resolution	3840×2160
Ego FOV Mode	string	Egocentric field-of-view setting	Documented per clip
Ego FPS	float	Egocentric frame rate	29.97 or 59.94 (per-clip)
Ego File Size (GB)	float	Egocentric file size	~1.5–20 GB
Exo Device	string	Exocentric camera model	Fixed-mount
Exo Resolution	string	Exocentric resolution	2688×1520 or 1920×1080
Exo FPS	float	Exocentric frame rate	30 (per-clip verified)
Exo File Size (GB)	float	Exocentric file size	~0.4–1.0 GB
Ego Sync Frame	integer	Ego frame index aligned to Exo Sync Frame	≥ 0
Exo Sync Frame	integer	Exo frame index aligned to Ego Sync Frame	≥ 0
Sync Lag (s)	float	Start-time offset between streams	Negative = ego started first
Sync Confidence	string	Confidence rating of sync estimate	High, Medium
Sync Method	string	Synchronization technique	Audio waveform correlation (8 kHz mono, 0.05 s RMS-delta envelope)
Ego File	string	Egocentric filename	Matches Clip ID convention
Exo File	string	Exocentric filename	Matches Clip ID convention

---

Additional notes: Sessions follow repeated procedural cycles (destroy → clean → reset) in fixed rooms with rotating object inventories, making the dataset well suited to skill-learning and procedural benchmarks. Delivered audio contains no third-party copyrighted music (cleanup sessions are captured music-free; destruction-session audio is delivered stripped or source-separated — specify preference at licensing). Directed-capture programs are available: licensees may specify camera configurations, object inventories, and task protocols for monthly ongoing delivery.An evaluation sample (manifest + representative clips) is available under a click-through evaluation agreement. Contact: corey@wreckvegas.com.