Research map / Vision-language and multimodal / Cross-modal learning
Found in Translation: Learning Robust Joint Representations by Cyclic Translations between Modalities
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 |
|---|---|---|---|
| Use language as source rather than visual or acoustic (ablation) | MCTN bimodal translation · CMU-MOSI | Acc: 53.1; 76.4; 76.8 | Helped |
| Add cyclic consistency to bimodal translation (ablation) | MCTN Bimodal · CMU-MOSI | Acc: 75.7 -> 76.8 | Helped |
| Use one shared Seq2Seq model instead of separate forward/backward models (ablation) | MCTN Bimodal · CMU-MOSI | Acc: 73.3 -> 75.2 | Helped |
| Add a second hierarchical modality translation level (ablation) | MCTN Trimodal · CMU-MOSI | Acc: 72.3 -> 79.3 | Helped |
| Add visual-to-acoustic cyclic representation before translating to language (ablation) | MCTN Trimodal · CMU-MOSI | Acc: 79.3 -> 56.4 | Hurt |
| Include language as the source modality (ablation) | MCTN Trimodal · CMU-MOSI | Acc: 56.4 -> 78.7; 79.3 | Helped |
| Add modalities during cyclic training (ablation) | MCTN · CMU-MOSI | Acc: 76.8 -> 79.3 | Helped |
| Use language as the only test-time input | MCTN · CMU-MOSI | Acc: 78.4 -> 79.3 | Helped |
| Use language as the only test-time input | MCTN · ICT-MMMO | Acc: 73.8 -> 81.3 | Helped |
| Use language as the only test-time input | MCTN · YouTube | F1: 51.6 -> 52.4 | Helped |
Closest papers by meaning
- Toward Robust Incomplete Multimodal Sentiment Analysis via Hierarchical Representation Learning (NeurIPS 2024 · similarity 0.90)
- SemMSA: Latent Semantic-Aided Robust Multimodal Sentiment Analysis with Incomplete Data (NeurIPS 2026 · similarity 0.88)
- DLF: Disentangled-Language-Focused Multimodal Sentiment Analysis (AAAI 2025 · similarity 0.88)
- Debiased Multimodal Understanding for Human Language Sequences (AAAI 2025 · similarity 0.88)
- Towards Robust Multimodal Sentiment Analysis with Incomplete Data (NeurIPS 2024 · similarity 0.87)
- A Unified Self-Distillation Framework for Multimodal Sentiment Analysis with Uncertain Missing Modalities (AAAI 2024 · 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.