Research map / Video / Video segmentation and tracking
Globally-Optimal Contrast Maximisation for Event Cameras
Xin Peng, Ling Gao, Yifu Wang, Laurent Kneip
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 |
|---|---|---|---|
| use global branch and bound instead of local optimisation (CMGD) | contrast maximisation optical flow · Optical flow (Line) | AEE: 1.12 -> 0.98 | Improved |
| use global branch and bound instead of local optimisation (CMGD) | contrast maximisation optical flow · Optical flow (Line) | runtime (s): 9.59 -> 28.73 | Hurt |
| use branch and bound instead of gradient ascent (ablation) | SoS contrast maximisation · Visual odometry Foam curve | RMS angular velocity error (deg/s): 11.5023 -> 3.0091 | Improved |
| use event-based GOCMF instead of frame-based GOVO | downward-facing planar motion estimation · Visual odometry Line | RMS angular velocity error (deg/s): 6.970 -> 0.517 | Improved |
| use GOCMF instead of local CMGD | rotational motion estimation · Rotational motion dynamic | mean error mu(epsilon): 168.7 -> 17.05 | Improved |
| use recursive tighter upper bound (GOCMF) instead of CMBNB bound (ablation) | branch and bound · Rotational motion dynamic | runtime: 76.46 -> 22.88 | Improved |
| use recursive bound (GOCMF) instead of CMBNB (ablation) | branch and bound · Rotational motion boxes | mu(epsilon): 30.21 -> 31.27 | Hurt |
| use GOCMF instead of CMBNB | branch and bound · Rotational motion poster | mu(epsilon): 32.94 -> 32.39 | No clear effect |
| downsample event stream by factor up to 6 (ablation) | GOCMF · boxes | error phi: 25 to 55 degrees | No clear effect |
| add salt and pepper noise with N/E ratio up to 0.4 (ablation) | GOCMF · synthetic lines | motion error | No clear effect |
| use GOCMF instead of local optimisation under downsampling | GOCMF · boxes | mean error phi: 30-250 degrees (CMGD) vs 25-55 degrees (GOCMF) | Improved |
Closest papers by meaning
- Globally-Optimal Event Camera Motion Estimation (ECCV 2020 · similarity 0.94)
- Globally Optimal Contrast Maximisation for Event-Based Motion Estimation (CVPR 2020 · similarity 0.92)
- A Unifying Contrast Maximization Framework for Event Cameras, With Applications to Motion, Depth, and Optical Flow Estimation (CVPR 2018 · similarity 0.91)
- Event Cameras, Contrast Maximization and Reward Functions: An Analysis (CVPR 2019 · similarity 0.91)
- Simultaneous Optical Flow and Intensity Estimation From an Event Camera (CVPR 2016 · similarity 0.90)
- Towards Robust Event-guided Low-Light Image Enhancement: A Large-Scale Real-World Event-Image Dataset and Novel Approach (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.
