5 papers
GeoUniPR: A Geometry-Consistent Unified Framework for Cross-Modal Place Recognition
Wonbong Kim, Jiatong Xiao, Rui Li +5
Cross-modal place recognition (CMPR) aims to identify the same location across heterogeneous sensing modalities, such as vision and LiDAR. Existing methods commonly bridge the moda…
Learning from Mistakes: Loss-Aware Memory Enhanced Continual Learning for LiDAR Place Recognition
Xufei Wang, Junqiao Zhao, Siyue Tao +3
LiDAR place recognition plays a crucial role in SLAM, robot navigation, and autonomous driving. However, existing LiDAR place recognition methods often struggle to adapt to new env…
Improving Sound Source Localization with Joint Slot Attention on Image and Audio
Inho Kim, Youngkil Song, Jicheol Park +2
Sound source localization (SSL) is the task of locating the source of sound within an image. Due to the lack of localization labels, the de facto standard in SSL has been to repres…
TailedCore: Few-Shot Sampling for Unsupervised Long-Tail Noisy Anomaly Detection
Yoon Gyo Jung, Jaewoo Park, Jaeho Yoon +4
We aim to solve unsupervised anomaly detection in a practical challenging environment where the normal dataset is both contaminated with defective regions and its product class dis…
Joint-Embedding Predictive Architecture for Self-Supervised Learning of Mask Classification Architecture
Dong-Hee Kim, Sungduk Cho, Hyeonwoo Cho +3
In this work, we introduce Mask-JEPA, a self-supervised learning framework tailored for mask classification architectures (MCA), to overcome the traditional constraints associated…