Research map / Video / Video segmentation and tracking
Focus Is All You Need: Loss Functions for Event-Based Vision
Guillermo Gallego, Mathias Gehrig, Davide Scaramuzza
Check 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 |
|---|---|---|---|
| Use IWE gradient magnitude instead of the timestamp-based loss | Event-camera rotational motion compensation · boxes_rotation | RMS angular velocity error (°/s): 82.89 -> 17.83 | Improved |
| Use IWE derivatives (gradient or Laplacian) rather than the variance baseline | Event-camera rotational motion compensation · boxes_rotation | RMS angular velocity error (°/s): 18.52 -> 17.58 | Improved |
| Use event polarity (ablation) | Focus-loss motion compensation · Boxes and Poster | RMS angular velocity error (°/s): 18.52 -> 18.94; 25.96 -> 24.39 | Mixed |
| Use variance rather than gradient magnitude | Focus-loss motion compensation · Boxes | runtime (µs): 128.46 -> 16.90 | Improved |
| Use timestamp-based alignment rather than event-count-based focus losses | Event-camera rotational motion compensation · boxes_rotation | RMS angular velocity error (°/s): 17.83 -> 82.89 | Hurt |
Closest papers by meaning
- Simultaneous Optical Flow and Intensity Estimation From an Event Camera (CVPR 2016 · similarity 0.89)
- Event-Based Vision: A Survey (TPAMI 2020 · similarity 0.89)
- Events-To-Video: Bringing Modern Computer Vision to Event Cameras (CVPR 2019 · similarity 0.88)
- Event Structural Valley: A Unified Theoretical and Practical Framework for Event Camera Autofocus (CVPR 2026 · similarity 0.88)
- Event-Based, 6-DOF Camera Tracking from Photometric Depth Maps (TPAMI 2017 · similarity 0.88)
- EvFocus: Learning to Reconstruct Sharp Images from Out-of-Focus Event Streams (ICML 2025 · 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.
