Challenge and Website Launch
TODO (around Jan 2026)
ECCV 2026 Workshop
Event-Based Multimodal Vision: Imaging, Perception, and Understanding
Imaging, Perception, and Understanding with Event-Based Multimodal Vision
Important Dates
Tentative timeline. Exact dates will be announced after ECCV 2026 workshop schedule confirmation.
Challenge Registration Deadline
TODO (around Mar 2026)
Final Submission Deadline
TODO (around Jun 2026)
Notification and Leaderboard Freeze
TODO (around Jul 2026)
ECCV 2026 Workshop
Sept 8-13, 2026 (exact workshop day: TODO)
About
Our workshop focuses on research at the intersection of event-based sensing and multimodal vision, spanning the full pipeline from sensing systems and low-level imaging to perception and high-level understanding.
Sensing Systems
- Event-based and neuromorphic vision with multimodal visual sensing.
- Hybrid systems with event cameras, RGB, LiDAR, IMU, and language.
- Datasets, simulators, benchmarks, and annotation pipelines.
- Spatial-temporal calibration, synchronization, and alignment.
Low-Level Imaging
- Event-guided illumination modeling and brightness adjustment.
- Enhancement under low-light, HDR, blur, noise, and reconstruction scenarios.
- Image or intensity reconstruction and physics-aware video synthesis.
- Illumination-aware representations and sensor-aware modeling.
Mid-Level Perception
- Optical flow, depth, visual odometry, SLAM, and 3D reconstruction.
- Segmentation, detection, feature tracking, and visual tracking.
- Multimodal fusion for challenging dynamic environments.
- Model-based, embedded, and learning-based event-driven perception.
High-Level Understanding
- Visual grounding, action understanding, and spatiotemporal reasoning.
- Event-based MLLMs and event-stream vision-language learning.
- Autonomous driving, robotics, and embodied intelligence.
- Cross-modal reasoning with asynchronous sparse event representations.
Computing and Hardware
- Novel event cameras, neuromorphic processors, and event-based systems.
- Efficient architectures, including state-space and lightweight designs.
- Sensor-aware learning grounded in event generation mechanisms.
Applications
- Computational imaging and industrial inspection.
- Autonomous driving, robotics, AR/VR, and edge intelligence.
- Deployment-oriented solutions for dynamic real-world scenarios.
Challenges
Each challenge now occupies one full row with a visual slot and complete link placeholders.
Low-Level Imaging Challenge
Event-guided image reconstruction and enhancement under extreme motion and difficult illumination, with multimodal fusion strategies.
Dataset: SEE 600K
Competition Page Dataset Link Baseline Code Evaluation Server Leaderboard
Mid-Level Perception Challenge
Benchmark multimodal event-based methods for detection, segmentation, and tracking in dynamic or low-light scenes.
Dataset: EBMV-Percept (placeholder)
Competition Page Dataset Link Baseline Code Evaluation Server Leaderboard
High-Level Understanding Challenge
Evaluate event-based multimodal reasoning and language-grounded understanding with interpretable outputs.
Dataset: EBMV-Understand (placeholder)
Competition Page Dataset Link Baseline Code Evaluation Server Leaderboard
Schedule
On-site agenda for workshop day. All slots below are placeholders and will be finalized later.
| Time | Session | Speaker / Notes |
|---|---|---|
| 08:30 - 09:00 | Registration | TODO |
| 09:00 - 09:15 | Opening and Welcome | TODO |
| 09:15 - 10:45 | Invited Talks Session I | TODO |
| 11:00 - 12:30 | Challenge Session | TODO |
| 14:00 - 15:30 | Invited Talks Session II | TODO |
| 15:45 - 16:45 | Poster and Discussion | TODO |
| 16:45 - 17:15 | Awards and Closing | TODO |
Invited Speakers
Speaker lineup and topical keywords are updated according to your latest list.
Kostas Daniilidis
UPenn, USA
Multi-view geometry, 3D scene reconstruction, and geometric vision theory.
Guillermo Gallego
TU Berlin, Germany
Event-based vision, continuous-time motion estimation, and spatiotemporal modeling.
Priyadarshini Panda
USC, USA
Neuromorphic computing, spiking neural networks, and event-driven learning systems.
Shintaro Shiba
Woven by Toyota, Japan
Event-based perception for autonomous driving and large-scale real-world deployment.
Liyuan Pan
BIT, China
Event cameras, spiking vision models, and high-speed visual perception.
Daniel Gehrig
UPenn, USA
Event-based optical flow, learning-based motion estimation, and high-speed vision.
Min Liu
DVSense, China
Event camera sensors, imaging hardware design, and industrial vision systems.
Jian Deng
AlpsenTek, Switzerland
Neuromorphic vision sensors, edge imaging devices, and low-power perception hardware.
Bo Mu
OmniVision, USA
Event-based vision applications, neuromorphic sensors, and real-world deployment.
Organizers
Organizer list and personal homepages.
Hanyu Zhou
National University of Singapore
Yunfan Lu
Hong Kong University of Science and Technology (Guangzhou)
Shaoyu Liu
Xidian University
Haoyue Liu
Huazhong University of Science and Technology
Peiqi Duan
Peking University
Shihan Peng
Huazhong University of Science and Technology
Yinqiang Zheng
University of Tokyo
Boxin Shi
Peking University
Kuk-Jin Yoon
KAIST
Hui Xiong
Hong Kong University of Science and Technology (Guangzhou)
Gim Hee Lee
National University of Singapore
Davide Scaramuzza
University of Zurich
