Research map / Detection and segmentation / Open-set and few-shot detection
Recurrent Multimodal Interaction for Referring Image Segmentation
Chenxi Liu, Zhe Lin, Xiaohui Shen, Jimei Yang, Xin Lu, Alan Yuille
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 |
|---|---|---|---|
| replace language-only LSTM with convolutional multimodal LSTM (RMI) (ablation) | ResNet model · Google-Ref val | IoU: 28.60 -> 32.06 | Improved |
| replace language-only LSTM with RMI (ablation) | DeepLab model · Google-Ref val | IoU: 33.08 -> 34.40 | Improved |
| replace language-only LSTM with RMI (ablation) | ResNet model · UNC val | IoU: 38.74 -> 39.74 | Improved |
| replace language-only LSTM with RMI (ablation) | ResNet model · UNC+ val | IoU: 26.25 -> 27.85 | Improved |
| replace language-only LSTM with RMI (ablation) | ResNet model · ReferItGame test | IoU: 54.01 -> 54.55 | Improved |
| replace language-only LSTM with RMI (ablation) | DeepLab model · ReferItGame test | IoU: 56.83 -> 57.34 | Improved |
| add DenseCRF refinement | RMI with DeepLab · UNC val | IoU: 44.33 -> 45.18 | Improved |
| add DenseCRF refinement | LSTM baseline with ResNet · Google-Ref val | IoU: 28.60 -> 28.94 | Improved |
| add DenseCRF refinement | RMI with ResNet · Google-Ref val | IoU: 32.06 -> 32.85 | Improved |
| replace ResNet weights with DeepLab weights | LSTM baseline · Google-Ref val | IoU: 28.60 -> 33.08 | Improved |
| replace ResNet weights with DeepLab weights | RMI · Google-Ref val | IoU: 32.06 -> 34.40 | Improved |
| use RMI instead of LSTM on long expressions (11-20 words) (ablation) | ResNet model · Google-Ref val (length 11-20) | IoU: 26.61 -> 30.56 | Improved |
Closest papers by meaning
- Weakly Supervised Referring Image Segmentation with Intra-Chunk and Inter-Chunk Consistency (ICCV 2023 · similarity 0.90)
- Segmentation from Natural Language Expressions (ECCV 2016 · similarity 0.90)
- Linguistic Structure Guided Context Modeling for Referring Image Segmentation (ECCV 2020 · similarity 0.90)
- Beyond One-to-One: Rethinking the Referring Image Segmentation (ICCV 2023 · similarity 0.90)
- See Say and Segment: Teaching LMMs to Overcome False Premises (CVPR 2024 · similarity 0.89)
- Prompt-Driven Referring Image Segmentation with Instance Contrasting (CVPR 2024 · similarity 0.89)
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.
