Research map / Video / Action recognition
Generating an Event Timeline About Daily Activities From a Semantic Concept Stream
Taiki Miyanishi, Jun-ichiro Hirayama, Takuya Maekawa, Motoaki Kawanabe
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add temporal relations model (ablation) | GRU-Lang · ADL event timeline, TrueConcept | Bleu@2: 0.6475 -> 0.6769 | Improved |
| add temporal relations model (ablation) | GRU-Lang · ADL event timeline, TrueConcept | CIDEr: 3.9751 -> 4.1519 | Improved |
| add temporal relations model (ablation) | GRU-Lang · ADL event timeline, PredConcept | Bleu@2: 0.4739 -> 0.5092 | Improved |
| add temporal relations model (ablation) | GRU-Lang · ADL event timeline, PredConcept | CIDEr: 2.6507 -> 2.8646 | Improved |
| add event sequence (context) model (ablation) | GRU-Lang+Time · ADL event timeline, TrueConcept | Bleu@3: 0.6279 -> 0.6768 | Improved |
| add event sequence (context) model (ablation) | GRU-Lang+Time · ADL event timeline, TrueConcept | CIDEr: 4.1519 -> 4.3306 | Improved |
| add event sequence (context) model (ablation) | GRU-Lang+Time · ADL event timeline, PredConcept | Bleu@3: 0.4276 -> 0.4776 | Improved |
| add event sequence (context) model (ablation) | GRU-Lang+Time · ADL event timeline, PredConcept | CIDEr: 2.8646 -> 3.1421 | Improved |
| GRU language model instead of bigram language model | event selection · ADL event timeline, TrueConcept | CIDEr: 3.8151 -> 3.9751 | Improved |
| GRU language model instead of bigram language model | event selection · ADL event timeline, PredConcept | Bleu@1: 0.6925 -> 0.7115 | No clear effect |
| external-corpus GRU language model instead of end-to-end S2VT | event selection · ADL event timeline, TrueConcept | Bleu@1: 0.0955 -> 0.8392 | Improved |
| use predicted instead of true semantic concepts | GRU-Lang+Time+Context · ADL event timeline | Bleu@1: 0.8546 -> 0.7450 | Hurt |
Closest papers by meaning
- Semantic Event Fusion of Different Visual Modality Concepts for Activity Recognition (TPAMI 2016 · similarity 0.85)
- Predicting Human Activities Using Stochastic Grammar (ICCV 2017 · similarity 0.84)
- RHYTHM: Reasoning with Hierarchical Temporal Tokenization for Human Mobility (NeurIPS 2025 · similarity 0.83)
- HARDVS: Revisiting Human Activity Recognition with Dynamic Vision Sensors (AAAI 2024 · similarity 0.82)
- SensorLM: Learning the Language of Wearable Sensors (NeurIPS 2025 · similarity 0.82)
- SNAP: Semantic Stories for Next Activity Prediction (AAAI 2025 · similarity 0.81)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
