File size: 1,692 Bytes
95a9764 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 | {
"term": "ActionRecognition",
"role": "concept",
"parent_concepts": [
"MultimodalReasoning"
],
"layer": 4,
"domain": "ComputerScience",
"definition": "Identifying human actions and activities in video by analyzing motion and temporal patterns",
"definition_source": "SUMO",
"aliases": [
"ActivityRecognition",
"ActionDetection",
"HumanActivityRecognition"
],
"wordnet": {
"synsets": [],
"canonical_synset": "",
"lemmas": [],
"pos": "noun"
},
"relationships": {
"related": [
"VideoUnderstanding",
"PoseEstimation",
"MotionAnalysis"
],
"antonyms": [],
"has_part": [],
"part_of": []
},
"safety_tags": {
"risk_level": "low",
"impacts": [],
"treaty_relevant": false,
"harness_relevant": false
},
"training_hints": {
"positive_examples": [
"Action recognition classified the video as 'playing basketball' with 95% confidence.",
"The model recognizes fine-grained actions like 'pouring' vs 'stirring' in cooking videos.",
"Temporal action detection locates when each action starts and ends in untrimmed video.",
"Two-stream networks combine appearance and optical flow for action recognition."
],
"negative_examples": [
"The person is doing something.",
"There is activity in the video.",
"Someone is moving."
],
"disambiguation": "Computational classification of actions from video, not general observation"
},
"is_category_lens": true,
"child_count": 0,
"meld_source": {
"meld_id": "org.hatcat/multimodal-fusion@0.1.0",
"applied_at": "2025-12-10T20:54:17.621328Z",
"pack_version": "5.7.3"
}
} |