An alternative open source neural network framework written in C and #CUDA | #deeplearning #sourcecode #argovision

https://github.com/pjreddie/darknet

GTC Europe by NVIDIA, 3-day conference on #AI, self-driving cars, #bigdata, #VR, etc. | #cuda #deeplearning

Full program: https://www.gputechconf.eu/Sessions.aspx?ncid=em-ded-mt-21286

🔥Neighborhood Attention Transformer 🔥

👉A novel transformer for both image classification and downstream vision tasks

𝐇𝐢𝐠𝐡𝐥𝐢𝐠𝐡𝐭𝐬:
✅Neighborhood Attention (NA)
✅Neighborhood Attention Transformer, NAT
✅Faster training/inference, good throughput
✅Checkpoints, train, #CUDA kernel available

More: https://bit.ly/3F5aVSo

🧨Fast-SNARF: 150× faster NeRF🧨

👉Voxel-based, pre-computed skinning, #cuda kernels -> avatar 150x faster!

😎Review https://bit.ly/3OMvEPo
😎Code github.com/xuchen-ethz/fast-snarf
😎Paper dataset.ait.ethz.ch/downloads/fast-snarf/paper.pdf