MobilitetsLab Stor-Trondheim

News of the week

IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2025

Conference Location: Tucson, Arizona, USA

Conference Dates: February 28 - March 4, 2025

Conference Focus: WACV is a premier international computer vision event focusing on practical applications and innovative algorithms in computer vision. The conference solicits high-quality, original submissions describing research in computer vision, with a particular emphasis on systems and applications.

Conference Key Areas:

The conference addresses various computer vision topics such as

- 3D computer vision

- Image recognition and understanding

- Machine learning architectures for vision

- Video recognition and understanding

It also addresses covers a wide range of application areas, including:

- Agriculture

- Autonomous driving

- Biomedical and healthcare

- Environmental monitoring

- Robotics

- Virtual and augmented reality

Sachin Verma's (MoST ph.d. candidate) experience :

Presenting our work, *"SegDesicNet: ..."*, at WACV 2025 was an exhilarating experience. In this work, I introduced a novel approach to data domain adaptation by bridging the gap between artificial neural networks and human brain-inspired modeling while incorporating the Earth's spherical geometry into my semantic segmentation model. This integration enhances the model with elements of explainable AI, improving interpretability and alignment with real-world spatial structures.

WACV is a premier venue for computer vision research, attracting top researchers from institutions like Google Brain, Meta, Stanford, and NVIDIA. The conference is highly competitive, with two rounds of submissions. I was fortunate to be accepted in Round 1, which saw 1,381 valid paper submissions, with only 167 accepted—an acceptance rate of just ~12%. Notably, WACV ranks 5th in computer vision and 9th as one of the most desirable venues for research submissions.

Engaging with some of the best minds in the field was truly inspiring. I had the opportunity to discuss cutting-edge challenges with researchers from major tech companies and leading institutions, many working on problems closely related to our research. These conversations led to insightful feedback and exciting possibilities for future collaboration. In particular, our discussions on scaling Gaussian Splats with geospatial AI for 3D reconstruction opened up promising new directions.

Beyond the technical takeaways, the atmosphere at WACV was electrifying—being surrounded by brilliant researchers, exchanging ideas, and witnessing the future of computer vision made this an unforgettable experience. This was also my second conference presentation—the first being at *NetZero*—and that experience played a crucial role in preparing me for WACV.

One of the most rewarding aspects was the unwavering support from both of my professors, Gabriel and Frank. Even amidst their busy schedules, they provided invaluable guidance, ensuring that our work reached its full potential. I am deeply grateful for their mentorship and encouragement. A huge thank you to the entire MoST research team for their support in making this possible!