Research map / Video / Video segmentation and tracking
Continuous Sign Language Recognition With Correlation Network
Read the paperCheck 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 CorrNet to the ResNet18 CSLR system (ablation) | ResNet18 CSLR · PHOENIX14 | WER: 21.0 -> 19.4 | Helped |
| add CorrNet to the ResNet18 CSLR system | ResNet18 CSLR · PHOENIX14-T | WER | Mixed |
| add CorrNet to the CSLR system | CSLR · CSL-Daily | WER: 32.0 -> 30.1 | Helped |
| add CorrNet to the CSLR system | CSLR · CSL | WER: 0.9 -> 0.8 | Helped |
| add the identification module with spatial dilation branches up to Ns=3 (ablation) | CorrNet · PHOENIX14 Test | WER: 21.0 -> 19.4 | Helped |
| increase spatial dilation branches from Ns=1 to Ns=3 (ablation) | Identification module · PHOENIX14 Test | WER: 20.1 -> 19.4 | Helped |
| increase spatial dilation branches from Ns=3 to Ns=4 (ablation) | Identification module · PHOENIX14 Test | WER: 19.4 -> 19.7 | Hurt |
| use temporal dilation branches Nt=4 rather than Nt=2, 3, or 5 (ablation) | Identification module · PHOENIX14 Test | WER: 19.9; 19.7; 19.8 -> 19.4 | Helped |
| replace multi-scale identification branches with a normal convolution (ablation) | CorrNet · PHOENIX14 Test | WER: 19.4 -> 20.4 | Hurt |
| expand the correlation neighborhood from K=3 to the full image (ablation) | Correlation module · PHOENIX14 Test | WER: 20.4 -> 19.4 | Helped |
| use correlation and identification modules together rather than either module alone (ablation) | CorrNet · PHOENIX14 Test | WER: 20.0; 19.9 -> 19.4 | Helped |
| place CorrNet after Stages 2, 3, and 4 rather than after one or two stages (ablation) | CorrNet · PHOENIX14 Test | WER: 20.1; 20.2; 20.0; 19.9 -> 19.4 | Helped |
Closest papers by meaning
- TCNet: Continuous Sign Language Recognition from Trajectories and Correlated Regions (AAAI 2024 · similarity 0.92)
- VSNet: Focusing on the Linguistic Characteristics of Sign Language (CVPR 2025 · similarity 0.87)
- SignGraph: A Sign Sequence is Worth Graphs of Nodes (CVPR 2024 · similarity 0.87)
- OLMD: Orientation-aware Long-term Motion Decoupling for Continuous Sign Language Recognition (AAAI 2025 · similarity 0.86)
- Discrete to Continuous: Generating Smooth Transition Poses from Sign Language Observations (CVPR 2025 · similarity 0.84)
- KD-MSLRT: Lightweight Sign Language Recognition Model Based on Mediapipe and 3D to 1D Knowledge Distillation (AAAI 2025 · similarity 0.84)
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.