4 papers
Copy-Pasting Coherent Depth Regions Improves Contrastive Learning for Urban-Scene Segmentation
Liang Zeng, Attila Lengyel, Nergis Tömen +1
In this work, we leverage estimated depth to boost self-supervised contrastive learning for segmentation of urban scenes, where unlabeled videos are readily available for training…
VIPriors 2: Visual Inductive Priors for Data-Efficient Deep Learning Challenges
Attila Lengyel, Robert-Jan Bruintjes, Marcos Baptista Rios +4
The second edition of the "VIPriors: Visual Inductive Priors for Data-Efficient Deep Learning" challenges featured five data-impaired challenges, where models are trained from scra…
Spectral Leakage and Rethinking the Kernel Size in CNNs
Nergis Tomen, Jan van Gemert
Convolutional layers in CNNs implement linear filters which decompose the input into different frequency bands. However, most modern architectures neglect standard principles of fi…
Top-Down Networks: A coarse-to-fine reimagination of CNNs
Ioannis Lelekas, Nergis Tomen, Silvia L. Pintea +1
Biological vision adopts a coarse-to-fine information processing pathway, from initial visual detection and binding of salient features of a visual scene, to the enhanced and prefe…