🇮🇹 Attending IEEE AVSS 2026

I attended the 22nd IEEE International Conference on Advanced Video and Signal-Based Surveillance (AVSS 2026), held in Lecce, Italy, under the theme “Beyond Monitoring”. The conference opened at the Convitto Palmieri and continued in the beautiful setting of the Carlo V Castle — a memorable venue for a few days of talks on video analysis, surveillance, and signal-based systems.

---

🎤 TraRA at AVSS 2026

Our paper, “TraRA: Trajectory-level Recognition Aggregation for Video Text Spotting in Urban Surveillance”, was presented by Duc Tri Tran, the first author of the work. Tri gave a great talk introducing the motivation behind trajectory-level aggregation for reading text in urban surveillance video, where individual frames are often blurry, occluded, or low-resolution, and where aggregating recognition results along a text trajectory leads to far more reliable spotting.

The work is a collaboration between RIKEN, Hanoi University of Science and Technology, Nagoya University, Lawrence Technological University, and Ritsumeikan University.

---

🤝 Conversations in Lecce

Beyond our own session, it was a pleasure to follow the talks and posters across the program and to discuss with researchers working on closely related problems in surveillance and video understanding. The discussions after the presentation were especially useful — several questions pointed directly at directions we are curious to explore next.

---

Gratitude

Many thanks to the AVSS 2026 organizers for a well-run conference in such a remarkable city, and congratulations to Tri on presenting his work. Thanks as well to all our collaborators for making this paper possible.