Deformable Kernels: Adapting Effective Receptive Fields for Object Deformation

Convolutional networks are not aware of an object's geometric variations, which leads to inefficient utilization of model and data capacity. To overcome this issue, recent works on deformation modeling seek to spatially reconfigure the data towards a common arrangement such that semantic recognition suffers less from deformation... (read more)

PDF Abstract ICLR 2020 PDF ICLR 2020 Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Object Detection COCO test-dev ResNet-50-DW-DPN (Deformable Kernels) box AP 40.6 # 59
APS 24.6 # 53
APM 43.9 # 60
APL 53.3 # 54
Image Classification ImageNet ResNet-50-DW (Deformable Kernels) Top 1 Accuracy 78.5% # 90

Methods used in the Paper