Research map / Video / Action recognition
Hollywood in Homes: Crowdsourcing Data Collection for Activity Understanding
Gunnar A. Sigurdsson, Gül Varol, Xiaolong Wang, Ali Farhadi, Ivan Laptev, Abhinav Gupta
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 |
|---|---|---|---|
| increase GMM vocabulary from 64 to 256 components (ablation) | IDT · Charades action classification | mAP (%): 16.5 -> 17.2 | Improved |
| combine HOG, HOF and MBH descriptors instead of HOG alone (ablation) | IDT (K=256) · Charades action classification | mAP (%): 13.0 -> 17.2 | Improved |
| add MBH to HOG descriptors (ablation) | IDT (K=256) · Charades action classification | mAP (%): 13.0 -> 16.5 | Improved |
| add late fusion of all baselines | Combined · Charades action classification | mAP (%): 17.2 -> 18.6 | Improved |
| add class-balanced sampling to two-stream training (ablation) | Two-Stream · Charades action classification | mAP (%): 14.3 -> 11.9 | Hurt |
| restrict evaluation to actions without a specific interacted object | Combined · Charades action classification | mAP (%): 18.6 -> 38.9 | Improved |
| predict video descriptions instead of scripts | S2VT · Charades sentence prediction | CIDEr: 0.17 -> 0.14 | Mixed |
| add sign-up bonuses for worker recruitment | Hollywood in Homes · Crowdsourced collection | new worker rate: 211% increase | Improved |
| add performance bonuses every 15th video | Hollywood in Homes · Crowdsourced collection | worker retention and output: 34% and 109% increases | Improved |
Closest papers by meaning
- AVA: A Video Dataset of Spatio-Temporally Localized Atomic Visual Actions (CVPR 2018 · similarity 0.87)
- About Time: Advances, Challenges, and Outlooks of Action Understanding (IJCV 2025 · similarity 0.87)
- What Actions Are Needed for Understanding Human Actions in Videos? (ICCV 2017 · similarity 0.86)
- Scenes-Objects-Actions: A Multi-Task, Multi-Label Video Dataset (ECCV 2018 · similarity 0.86)
- The EPIC-KITCHENS Dataset: Collection, Challenges and Baselines (TPAMI 2020 · similarity 0.86)
- Moments in Time Dataset: One Million Videos for Event Understanding (TPAMI 2019 · similarity 0.86)
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.
