Research map / Video / Video generation
MonSTeR: a Unified Model for Motion, Scene, Text Retrieval
Luca Collorone, Matteo Gioia, Massimiliano Pappa, Paolo Leoni, Giovanni Ficarra, Or Litany, Indro Spinelli, Fabio Galasso
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 cross-modal encoders and cross-modal contrastive terms (ablation) | MonSTeR · HUMANISE+ (All protocol) | avg mRecall (%): 3.79 -> 5.63 | Improved |
| add cross-modal encoders and cross-modal contrastive terms (ablation) | MonSTeR · HUMANISE+ (Small Batches protocol) | avg mRecall (%): 56.07 -> 60.00 | Improved |
| add single-modality contrastive terms (ablation) | MonSTeR · HUMANISE+ (All protocol) | avg mRecall (%): 4.36 -> 5.63 | Improved |
| add single-modality contrastive terms (ablation) | MonSTeR · HUMANISE+ (Small Batches protocol) | avg mRecall (%): 41.77 -> 60.00 | Improved |
| add single-modality contrastive terms (ablation) | MonSTeR · TRUMANS+ (All protocol) | avg mRecall (%): 5.34 -> 7.42 | Improved |
| use separate unimodal and cross-modal encoders instead of one tri-modal encoder (ablation) | MonSTeR · HUMANISE+ (All protocol) | avg mRecall (%): 4.14 -> 5.63 | Improved |
| use separate unimodal and cross-modal encoders instead of one tri-modal encoder (ablation) | MonSTeR · TRUMANS+ (Small Batches protocol) | avg mRecall (%): 46.72 -> 47.04 | Improved |
| add cross-modal encoders (ablation) | MonSTeR · TRUMANS+ (All protocol) | avg mRecall (%): 6.27 -> 7.42 | Improved |
| use MonSTeR embeddings in GPT2 captioner instead of MotionGPT | MonSTeR + GPT2 · HUMANISE+ captioning | BLEU 4: 17.47 -> 23.59 | Improved |
| use MonSTeR embeddings in GPT2 captioner instead of MotionGPT | MonSTeR + GPT2 · TRUMANS+ captioning | CIDEr: 15.08 -> 12.24 | Hurt |
Closest papers by meaning
- Chronologically Accurate Retrieval for Temporal Grounding of Motion-Language Models (ECCV 2024 · similarity 0.90)
- ReMoGPT: Part-Level Retrieval-Augmented Motion-Language Models (AAAI 2025 · similarity 0.89)
- ReMoMask: Retrieval-Augmented Masked Motion Generation (ECCV 2026 · similarity 0.89)
- VimoRAG: Video-based Retrieval-augmented 3D Motion Generation for Motion Language Models (NeurIPS 2025 · similarity 0.87)
- LaMP: Language-Motion Pretraining for Motion Generation, Retrieval, and Captioning (ICLR 2025 · similarity 0.87)
- Tri-Modal Motion Retrieval by Learning a Joint Embedding Space (CVPR 2024 · similarity 0.87)
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.
