Research map / Video / Video-language understanding
Video Captioning With Transferred Semantic Attributes
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 semantic attributes from images and videos to LSTM | LSTM video captioning · MSVD | CIDEr-D: 68.0 -> 74.0 | Helped |
| add semantic attributes from images and videos to LSTM | LSTM video captioning · M-VAD | METEOR: 6.8 -> 7.2 | Helped |
| add semantic attributes from images and videos to LSTM | LSTM video captioning · MPII-MD | METEOR: 7.3 -> 8.0 | Helped |
| add image attributes only (ablation) | LSTM video captioning · MSVD | CIDEr-D | Helped |
| add video attributes only (ablation) | LSTM video captioning · MSVD | CIDEr-D | Helped |
| combine image and video attributes rather than use video attributes alone (ablation) | LSTM-TSA_V · MSVD | CIDEr-D: 71.7 -> 74.0 | Helped |
| learn video attributes jointly across all sampled frames with video MIL instead of frame-by-frame image MIL (ablation) | LSTM-TSA_V · MSVD | CIDEr-D: 70.6 -> 71.7 | Helped |
| replace a fixed sum with a contextual transfer gate that fuses both attribute sources (ablation) | LSTM-TSA_IV · MSVD | CIDEr-D: 71.7 -> 74.0 | Helped |
| gate both image and video attributes with decoupled weights rather than gate image attributes only (ablation) | LSTM-TSA_IV · MSVD | CIDEr-D: 71.5 -> 74.0 | Helped |
| gate both image and video attributes with decoupled weights rather than gate video attributes only (ablation) | LSTM-TSA_IV · MSVD | CIDEr-D: 72.3 -> 74.0 | Helped |
Closest papers by meaning
- Fully Convolutional Video Captioning with Coarse-to-Fine and Inherited Attention (AAAI 2019 · similarity 0.89)
- Jointly Modeling Embedding and Translation to Bridge Video and Language (CVPR 2016 · similarity 0.89)
- Reference Based LSTM for Image Captioning (AAAI 2017 · similarity 0.88)
- Convolutional Image Captioning (CVPR 2018 · similarity 0.88)
- Incorporating Copying Mechanism in Image Captioning for Learning Novel Objects (CVPR 2017 · similarity 0.88)
- Temporal Deformable Convolutional Encoder-Decoder Networks for Video Captioning (AAAI 2019 · similarity 0.88)
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.