RF-Next: Efficient Receptive Field Search for Convolutional Neural Networks. (arXiv:2206.06637v2 [cs.CV] UPDATED)

Web: http://arxiv.org/abs/2206.06637

Temporal/spatial receptive fields of models play an important role in
sequential/spatial tasks. Large receptive fields facilitate long-term
relations, while small receptive fields help to capture the local details.
Existing methods construct models with hand-designed receptive fields in
layers. Can we effectively search for receptive field combinations to replace
hand-designed patterns? To answer this question, we propose to find better
receptive field combinations through a global-to-local search scheme. Our
search scheme exploits both global search to find the coarse combinations and …

arxiv convolutional neural networks cv networks neural neural networks search