domain-video
Skill8 starsChanged 4 months agoMITView on GitHub
Use for video understanding: action or activity recognition, video classification, temporal action detection or localization in untrimmed video, gesture recognition, and general video analysis. Picks the right video model and the correct video-level split and metric. Triggers on 'video classification', 'action recognition', 'activity recognition', 'temporal action detection', 'video understanding', 'recognize actions', 'gesture', 'video dataset'.
In this repository22 other files it gives its agents.
- accuracy-improvement-loopSkill
- data-rigor-and-leakageSkill
- dataset-profilingSkill
- deployment-explainabilitySkill
- domain-3dSkill
- domain-anomaly-detectionSkill
- domain-audio-speechSkill
- domain-computer-visionSkill
- domain-generativeSkill
- domain-graphSkill
- domain-medical-imagingSkill
- domain-multimodalSkill
- domain-nlp-llmSkill
- domain-recommenderSkill
- domain-reinforcement-learningSkill
- domain-tabularSkill
- domain-time-seriesSkill
- literature-reviewSkill
- ml-research-methodologySkill
- notebook-deliverySkill
- rigorous-evaluationSkill
- training-optimizationSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Prefer models pretrained on large video datasets (Kinetics).
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Video Understanding - Method Selection
Prefer models pretrained on large video datasets (Kinetics). Video is compute-heavy, so use mixed precision and clip-based training.
Decision table
| Task | Recommended (recent, 2021-2025) | Notes |
| Action or video classification | VideoMAE or VideoMAEv2 (self-supervised pretrain, then fine-tune), Video Swin, UniFormerV2 | Kinetics-pretrained. Metric is top-1 and top-5 accuracy. |
| Efficient or low-budget | adapt a strong image backbone with temporal modules (AIM), or a small MoViNet | cheaper than a full video transformer. |
| Temporal action detection (localize actions in long untrimmed video) | ActionFormer, TadTR | metric is mAP at temporal IoU. |
| General-purpose video features | InternVideo family | strong for many downstream tasks. |
Cross-cutting practice
- Sample frames (uniform or dense), train on short clips, and average multiple clips at test time.
- Augment with temporal cropping, per-frame flip and light photometric jitter. Keep flips label-safe.
- Leakage: split by VIDEO or subject, never by frame or by clips taken from the same video.
- Metrics: top-1 and top-5 accuracy for trimmed classification; mAP at temporal IoU for detection. Not frame accuracy.
- Explainability: per-frame Grad-CAM or spatiotemporal attention; show which frames the model attended to.
- Improve accuracy: use
accuracy-improvement-loop; evaluate withrigorous-evaluation.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.