Research map / Time series / Foundation models for time series
FOCAL: Contrastive Learning for Multimodal Time-Series Sensing Signals in Factorized Orthogonal Latent Space
Shengzhong Liu, Tomoyoshi Kimura, Dongxin Liu, Ruijie Wang, Jinyang Li, Suhas Diggavi, Mani Srivastava, Tarek Abdelzaher
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 |
|---|---|---|---|
| remove private space and its contrastive task (ablation) | FOCAL (SW-T, linear) · MOD | accuracy: 0.9296 -> 0.9805 | Improved |
| remove private space and its contrastive task (ablation) | FOCAL (SW-T, linear) · ACIDS | accuracy: 0.7981 -> 0.9489 | Improved |
| remove orthogonality constraint (ablation) | FOCAL (SW-T, linear) · ACIDS | accuracy: 0.9311 -> 0.9489 | Improved |
| remove orthogonality constraint (ablation) | FOCAL (SW-T, linear) · RealWorld-HAR | accuracy: 0.9186 -> 0.9451 | Improved |
| replace orthogonality with distributional independence (ablation) | FOCAL (SW-T, linear) · MOD | accuracy: 0.9805 -> 0.5773 | Hurt |
| replace orthogonality with distributional independence (ablation) | FOCAL (SW-T, linear) · ACIDS | accuracy: 0.9489 -> 0.4926 | Hurt |
| remove temporal structural constraint (ablation) | FOCAL (SW-T, linear) · PAMAP2 | accuracy: 0.8367 -> 0.8580 | Improved |
| remove temporal structural constraint (ablation) | FOCAL (SW-T, linear) · RealWorld-HAR | accuracy: 0.9361 -> 0.9451 | Improved |
| replace temporal constraint with temporal contrastive task (ablation) | FOCAL (SW-T, linear) · MOD | accuracy: 0.9805 -> 0.9363 | Hurt |
| replace temporal constraint with temporal contrastive task (ablation) | FOCAL (SW-T, linear) · PAMAP2 | accuracy: 0.8580 -> 0.8391 | No clear effect |
| add temporal constraint as plugin | SimCLR · ACIDS | accuracy: 0.7438 -> 0.7461 | Improved |
| add temporal constraint as plugin | Cocoa · ACIDS | accuracy: 0.6644 -> 0.8543 | Improved |
Closest papers by meaning
- MAESTRO : Adaptive Sparse Attention and Robust Learning for Multimodal Dynamic Time Series (NeurIPS 2025 · similarity 0.86)
- What to align in multimodal contrastive learning? (ICLR 2025 · similarity 0.86)
- Continual Multimodal Contrastive Learning (NeurIPS 2025 · similarity 0.85)
- Connect, Collapse, Corrupt: Learning Cross-Modal Tasks with Uni-Modal Data (ICLR 2024 · similarity 0.85)
- VCR: Learning Valid Contextual Representation for Incomplete Wearable Signals (NeurIPS 2026 · similarity 0.85)
- Multi-modal contrastive learning adapts to intrinsic dimensions of shared latent variables (NeurIPS 2025 · similarity 0.85)
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.
