Skip to content

Releases: xw-hu/Unveiling-Deep-Shadows

Weights_VideoShadowDetection

06 Sep 11:33
ad3e6e8
Compare
Choose a tag to compare

During training, ShadowSAM uses the size of 1024x1024, while others adopt the size of 512x512.

Weights_ShadowRemoval

06 Sep 13:14
ad3e6e8
Compare
Choose a tag to compare

_256: Input Size 256x256 during training
_512: Input Size 512x512 during training

Weights_ShadowDetection

06 Sep 12:50
ad3e6e8
Compare
Choose a tag to compare

_256: Input Size 256x256 during training
_512: Input Size 512x512 during training

Weights_InstanceShadowDetection

06 Sep 14:22
ad3e6e8
Compare
Choose a tag to compare
Weights_ISD

Update README.md

Weights_DocumentShadowRemoval

06 Sep 14:32
ad3e6e8
Compare
Choose a tag to compare

Input Size 512x512 during training

Results

05 Sep 11:53
de7679f
Compare
Choose a tag to compare

Qualitative results