Research map / Video / Video segmentation and tracking
Unsupervised Contour Tracking of Live Cells by Mechanical and Cycle Consistency Losses
Junbong Jang, Kwonmoo Lee, Tae-Kyun Kim
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 mechanical-normal loss to cycle consistency loss (ablation) | contour tracker · Phase contrast live cell videos | SA_0.02: 0.632 -> 0.729 | Improved |
| Add mechanical-linear loss to the cycle+normal combination (ablation) | contour tracker · Phase contrast live cell videos | SA_0.02: 0.729 -> 0.426 | Hurt |
| Use photometric loss instead of geometric losses (ablation) | contour tracker · Phase contrast live cell videos | SA_0.02: 0.198 vs 0.632 (cycle only) | Hurt |
| Replace unsupervised losses with supervised point-to-point loss using mechanical-model correspondences (ablation) | contour tracker · Phase contrast live cell videos | SA_0.02: 0.549 vs 0.729 (full) | Hurt |
| Replace forward+backward cross attention with no cross attention (ablation) | contour tracker · Phase contrast live cell videos | SA_0.02: 0.659 vs 0.729 | Hurt |
| Use single cross attention instead of forward and backward cross attentions (ablation) | contour tracker · Phase contrast live cell videos | SA_0.02: 0.677 vs 0.729 | Hurt |
| Replace FCNN with circular convolution (ablation) | contour tracker · Phase contrast live cell videos | SA_0.02: 0.643 vs 0.729 | Hurt |
| Replace FCNN with 1D convolution (ablation) | contour tracker · Phase contrast live cell videos | SA_0.02: 0.692 vs 0.729 | Hurt |
| Proposed contour tracker vs mechanical model baseline | contour tracker · Phase contrast live cell videos | SA_0.02: 0.729 vs 0.683 | Improved |
| Proposed contour tracker vs mechanical model baseline | contour tracker · Confocal fluorescence live cell videos | SA_0.02: 0.632 vs 0.614 (PoST best) | Improved |
Closest papers by meaning
- Higher-Order Cell Tracking Transformer (NeurIPS 2026 · similarity 0.86)
- MorphoHELM: A Comprehensive Benchmark for Evaluating Representations for Microscopy-Based Morphology Assays (NeurIPS 2026 · similarity 0.83)
- Joint Velocity-Growth Flow Matching for Single-Cell Dynamics Modeling (NeurIPS 2025 · similarity 0.83)
- Beyond ViT Tokens: Masked-Diffusion Pretrained Convolutional Pathology Foundation Model for Cell-Level Dense Prediction (NeurIPS 2026 · similarity 0.82)
- CellFlux: Simulating Cellular Morphology Changes via Flow Matching (ICML 2025 · similarity 0.82)
- Towards Scalable Context-Aware Single-Cell Spatial Transcriptomics Prediction from Histology Images (NeurIPS 2026 · similarity 0.82)
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.
