Papers of the day   All papers

Gated-SCNN: Gated Shape CNNs for Semantic Segmentation

Comments

Aug 26 2019 Sanja Fidler

We now released code and trained models for our GSCNN semantic segmentation work done at @NvidiaAI. Check it out! @yongyuanxi @davidjesusacu @jampani_varun Paper (ICCV'19): https://arxiv.org/abs/1907.05740 Project page: https://nv-tlabs.github.io/GSCNN/ Code (@PyTorch ): https://github.com/nv-tlabs/gscnn https://t.co/kM0wTowcV3
2 replies, 635 likes


Jul 16 2019 Sanja Fidler

Excited to share our @NvidiaAI work on semantic segmentation called GSCNN, which significantly outperforms DeepLabV3+ on Cityscapes benchmark. @yongyuanxi @davidjesusacu @jampani_varun @NvidiaAI paper: https://arxiv.org/abs/1907.05740 project: https://github.com/nv-tlabs/GSCNN Code coming soon https://t.co/AKOlujDF5i
4 replies, 551 likes


Aug 26 2019 David Acuna

Training code in @PyTorch is now available for our latest ICCV 2019 state-of-the-art semantic segmentation network (G-SCNN). Check it out! https://github.com/nv-tlabs/gscnn @yongyuanxi @jampani_varun @FidlerSanja @NvidiaAI
0 replies, 49 likes


Jul 16 2019 David Acuna

Super excited of releasing G-SCNN A state-of-the-art semantic segmentation network that wires shape information into higher-level activations and exploits the duality between edges and sem segm. Great work with @yongyuanxi @jampani_varun @FidlerSanja @NvidiaAI
0 replies, 23 likes


Jul 16 2019 Andrey Kurenkov 🤖

"Specifically, we use the higher-level activations in the classical stream to gate the lower-level activations in the shape stream, effectively removing noise and helping the shape stream to only focus on processing the relevant boundary-related information." Neat!
0 replies, 7 likes


Jul 22 2019 HotComputerScience

Most popular computer science paper of the day: "Gated-SCNN: Gated Shape CNNs for Semantic Segmentation" https://hotcomputerscience.com/paper/gated-scnn-gated-shape-cnns-for-semantic-segmentation https://twitter.com/FidlerSanja/status/1151123795308089351
0 replies, 4 likes


Jul 15 2019 cs.LG Papers

Gated-SCNN: Gated Shape CNNs for Semantic Segmentation. Towaki Takikawa, David Acuna, Varun Jampani, and Sanja Fidler http://arxiv.org/abs/1907.05740
1 replies, 4 likes


Jul 15 2019 arXiv CS-CV

Gated-SCNN: Gated Shape CNNs for Semantic Segmentation http://arxiv.org/abs/1907.05740
0 replies, 3 likes


Aug 27 2019 高橋かずひと@リベロ拝命🐤

Amazing🐤
0 replies, 3 likes


Jul 16 2019 Sanja Fidler

Excited to share our @NvidiaA work on semantic segmentation called GSCNN, which significantly outperforms DeepLabV3+ on Cityscapes benchmark. @yongyuanxi @davidjesusacu @jampani_varun @NvidiaA paper: https://arxiv.org/abs/1907.05740 project page: https://github.com/nv-tlabs/GSCNN Code coming soon https://t.co/0ATD2F4Wkk
0 replies, 1 likes


Jul 15 2019 Brundage Bot

Gated-SCNN: Gated Shape CNNs for Semantic Segmentation. Towaki Takikawa, David Acuna, Varun Jampani, and Sanja Fidler http://arxiv.org/abs/1907.05740
1 replies, 0 likes


Content