Research map / Video / Action recognition
Summarizing First-Person Videos from Third Persons' Points of View
HSUAN-I HO, Wei-Chen Chiu, Yu-Chiang Frank Wang
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 unlabeled first-person videos (semi-supervised) (ablation) | Ours non-sequential · SumMe | F-measure: 35.272 -> 38.649 | Improved |
| add unlabeled first-person videos (semi-supervised) (ablation) | Ours non-sequential · FPVSum | F-measure: 37.098 -> 38.409 | Improved |
| add unlabeled first-person videos (semi-supervised) (ablation) | Ours sequential · SumMe | F-measure: 35.980 -> 41.991 | Improved |
| add unlabeled first-person videos (semi-supervised) (ablation) | Ours sequential · FPVSum | F-measure: 37.366 -> 38.572 | Improved |
| use recurrent highlight detector instead of 2-layer fully connected | Ours · SumMe | F-measure: 38.649 -> 41.991 | Improved |
| use recurrent highlight detector instead of 2-layer fully connected | Ours · FPVSum | F-measure: 38.409 -> 38.572 | Improved |
| add private encoders to shared encoder (ablation) | Ours non-sequential · SumMe | F-measure: 29.754 -> 38.649 | Improved |
| add private encoders to shared encoder (ablation) | Ours non-sequential · FPVSum | F-measure: 31.020 -> 38.409 | Improved |
| add pseudo triplet loss on unlabeled videos (ablation) | Ours non-sequential · SumMe | F-measure: 34.252 -> 38.649 | Improved |
| add pseudo triplet loss on unlabeled videos (ablation) | Ours non-sequential · FPVSum | F-measure: 35.485 -> 38.409 | Improved |
| use classification scoring loss instead of ranking loss (ablation) | Ours non-sequential · SumMe | F-measure: 28.435 -> 38.649 | Improved |
| use classification scoring loss instead of MSE loss (ablation) | Ours non-sequential · FPVSum | F-measure: 36.046 -> 38.409 | Improved |
Closest papers by meaning
- Highlight Detection With Pairwise Deep Ranking for First-Person Video Summarization (CVPR 2016 · similarity 0.90)
- First- And Third-Person Video Co-Analysis By Learning Spatial-Temporal Joint Attention (TPAMI 2020 · similarity 0.88)
- Generating Personalized Summaries of Day Long Egocentric Videos (TPAMI 2021 · similarity 0.86)
- Identifying First-Person Camera Wearers in Third-Person Videos (CVPR 2017 · similarity 0.86)
- Weakly Supervised Summarization of Web Videos (ICCV 2017 · similarity 0.86)
- Unsupervised Video Summarization With Adversarial LSTM Networks (CVPR 2017 · similarity 0.85)
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.
