CUSTOMVIDEOFIRST-PERSON
First-Person Activity Clips
Wearable-style first-person activity clips with task labels, step metadata, camera perspective, hands-visible review, environment fields, and optional narration for embodied AI evaluation.
CUSTOMVIDEOENVIRONMENT DIVERSITY
Multi-Environment Activity Clips
Activity video collection format for scene understanding, activity recognition, embodied AI generalization, and environment-robust model evaluation.
Multi-Environment Activity Clips are built for teams evaluating whether vision and multimodal models generalize across different real-world settings. The format captures similar or related activities across multiple environments so buyers can analyze how lighting, background, surface type, camera position, and environment context affect model behavior.
The dataset format is designed for buyers who need environment diversity without turning the collection into unstructured video footage.
Structured collection
Multi-Environment Activity Clips use defined activity tasks collected across varied settings.
Metadata depth
The format supports alignment between video timestamps, activity labels, step labels where scoped, object context, and…
Delivery-ready package
Includes video quality, activity match, environment metadata reviews, plus dataset manifest, CSV catalog, consent status…
Request sample access
Share your target activities, environment types, camera setup, annotation needs, and evaluation goals. We’ll reply with a technical scope for the collection.
Key highlights
- Activity clips organized by environment type, activity category, camera position, and lighting condition.
- Environment metadata prepared for filtering, robustness testing, and model performance analysis.
- Support for kitchen, workshop, garage, office, outdoor, retail, or buyer-scoped environment categories.
Dataset contents
Video capture
Video activity clips · Activity category labels · Environment type metadata · Environment sub-type metadata · Camera position metadata
Steps, transcripts, and labels
Lighting condition metadata · Duration metadata · Resolution and fps metadata · Contributor country or region where scoped · Task completion status
Capture and task metadata
Optional object context metadata · Optional step labels · Optional action labels
Technical specifications
Use cases
- Activity recognition
- Scene understanding
- Environment robustness testing
- Embodied AI generalization
- Vision-language grounding
- Multimodal model evaluation
- Context-shift analysis
- Real-world perception benchmarking
Dataset workflow
1Collection
Multi-Environment Activity Clips use defined activity tasks collected across varied settings. The structure is designed to separate the action being performed from the environment where it happens, so buyers can evaluate model robustness across context shifts.
2Review
Video quality review · Activity match review · Environment metadata review · Camera position review · Annotation consistency review · Consent status review · Delivery approval
3Delivery
MP4 video clips · Activity metadata CSV · Environment metadata CSV · Annotation files where scoped · Dataset manifest JSON · CSV file catalog · QA summary · Consent status metadata · Delivery notes
Quality assurance
Related dataset formats
CUSTOMVIDEODEXTEROUS TASKS
Hand and Object Interaction Clips
Hand-object interaction clips with action verbs, object categories, camera angle metadata, contact-event markers, object-state labels, and QA fields for robotic manipulation evaluation.
CUSTOMVIDEOINSTRUCTIONAL
Procedural Demonstration Video Sets
Step-structured procedural videos with task labels, timestamped step markers, narration alignment, completion status, and metadata for instruction-following and imitation learning.
CUSTOMAUDIOACOUSTIC SCENES
Acoustic Environment Sets
Ambient audio sets with scene labels, acoustic context metadata, speech screening, device fields, noise-level estimates, and QA status for acoustic scene and robustness evaluation.
CUSTOMMULTIMODALPAIRED DATA
Multimodal Paired Capture Sets
Paired multimodal datasets linking video, audio, screen, text, or image-derived records through shared file IDs, cross-modal manifests, metadata schemas, and QA fields.
CUSTOMANY MODALITYBUYER-SCOPED
Buyer-Scoped Custom Collection
Fully scoped custom data collection with buyer-defined modality, schema, metadata fields, annotation layers, QA criteria, manifest structure, delivery package, and collection rules.