One paper from NTT Laboratories have been accepted for publication for BMVC 2025

One paper authored by NTT Laboratories has been accepted for publication at BMVC 2025 (The 36th British Machine Vision Conference)(https://bmvc2025.bmva.org/), the flagship conference on computer vision to be held in Sheffield, UK, from 24th to 27th November, 2025. BMVC is known as one of the most competitive conferences in computer vision. The acceptance rate of BMVC 2025 is 31.9% (276 out of 865).

Abbreviated names of the laboratories:
HI: NTT Human Informatics Laboratories

Msmvd: Exploiting Multi-Scale Image Features via Multi-Scale Bev Features for Multi-View Pedestrian Detection

Taiga Yamane (HI), Satoshi Suzuki (HI), Ryo Masumura (HI), Shota Orihashi (HI), Tomohiro Tanaka (HI), Mana Ihori (HI), Naoki Makishima (HI), Naotaka Kawata (HI)

Multi-view pedestrian detection aims to detect pedestrians in the form of a bird's eye view from a multi-view image. In this task, deep learning methods have made significant progress. However, previous methods struggle to detect pedestrians with vastly large or small sizes in images. This study proposes a novel method that overcomes the limitations by utilising multi-scale image features capable of effectively representing pedestrians of diverse sizes.