ACCV 2026 Workshop on Advances in Event-based Computer Vision, December 15, 2026, Osaka, Japan

<aside> 📅

Tentative Program & Workshop Date & Time

0900-1230, December 15, 2026, Room 1006 (In-person in Osaka, Japan) — held in conjunction with ACCV 2026.

Each non-keynote presentation will have both an oral presentation and a poster presentation.

Program and presentation details are tentative and subject to change.

</aside>

<aside> 🧭

Contents

</aside>

Event-based cameras (neuromorphic sensors) represent a fundamental paradigm shift in computer vision. Unlike conventional frame-based cameras that capture static scenes at fixed intervals, event cameras respond asynchronously to per-pixel illumination changes with microsecond temporal resolution, exceptional high dynamic range (HDR > 120 dB), and minimal power consumption.

The demand for real-time computer vision in challenging, dynamic environments is expanding rapidly across applications like autonomous driving under extreme lighting, ultra-high-speed robotics, high-dynamic sports analysis, next-generation human-computer interaction (HCI), and resource-constrained edge computing. Traditional frame-based approaches face severe bottlenecks in these domains due to motion blur, high latency, and massive data redundancy.

AECV 2026 aims to bridge the gap between neuromorphic hardware and state-of-the-art CV algorithms, serving as a premier venue for discussions around modern deep learning foundations and real-world multi-modal fusion. This will be the inaugural event-based vision workshop at ACCV, strategically hosted in Japan — a country playing a pioneering role in the global semiconductor and advanced imaging sensor industries.

Topics of Interest

Topics of interest include, but are not limited to:

• Low-latency and high-speed algorithms for novel event-based application systems (e.g., autonomous driving, high-speed robotics, drones, sports analysis, and human-computer interactions)
• Exploration of novel sensing modalities and unconventional targets utilizing event-based vision (e.g., microscopic imaging, fluid dynamics, vibration analysis, and biomedical sensing)
• Event-based 3D reconstruction, SLAM, and 4D dynamic tracking
• High-speed object detection, tracking, and action recognition
• Bio-inspired vision and Spiking Neural Networks (SNNs)
• Multi-modal sensor fusion (Event + RGB, Event + LiDAR/IMU)
• Event-based image restoration, frame synthesis, and de-blurring
• Computational photography and advanced optics with neuromorphic sensors
• Edge-AI and low-power hardware deployment for event data
• New benchmarks, large-scale datasets, and high-fidelity simulators

Submission Tracks

AECV 2026 implements a structured three-track submission system to accommodate diverse types of research contributions: