Search Results for author:
Found 3 papers, 3 papers with code
TAO-Amodal: A Benchmark for Tracking Any Object Amodally
Amodal perception, the ability to comprehend complete object structures from partial visibility, is a fundamental skill, even for infants.
Self-Supervised Pyramid Representation Learning for Multi-Label Visual Analysis and Beyond
In particular, we present a cross-scale patch-level correlation learning in SS-PRL, which allows the model to aggregate and associate information learned across patch scales.
C3-SL: Circular Convolution-Based Batch-Wise Compression for Communication-Efficient Split Learning
Based on the simulation results on CIFAR-10 and CIFAR-100, our method achieves a 16x compression ratio with negligible accuracy drops compared with the vanilla SL.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
