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"
  }
}