20220124-20220206.md

意见性分享

---

课程和报告分享

• IJCV Special Issue

  Michael Black

  Computer vision for tracking and modeling animals -- IJCV Special Issue call for papers. Deadline March 27. We will collect state of the art papers in this exciting, growing, and impactful area. Please submit! https://springer.com/journal/11263/updates/19611514

  

  reference: https://twitter.com/Michael_J_Black/status/1487108610366885896?s=20&t=3nR1lbHi7T01m-xqQWUpkw

---

成果推荐及讨论

• Real Time Height Map Fusion

  Andrew Davison

  Real Time Height Map Fusion using Differentiable Rendering, with @jz4411 @StefanLeuteneg1, Dyson Robotics Lab, single RGB camera. Here used for dense, geometric drivable ground segmentation at <1cm height. (no learning needed).

  https://imperial.ac.uk/media/imperial-college/research-centres-and-groups/dyson-robotics-lab/jzienkiewicz_etal_iros2016.pdf

  CUDA for fast parallel inference. IROS 2016. #SpatialAI.

  reference: https://twitter.com/AjdDavison/status/1484810071498530818?s=20

• 3D scene reconstruction from video

  Vittorio Ferrari

  Our work on 3D scene reconstruction from video was accepted to #TPAMI!

  Vid2CAD: CAD Model Alignment using Multi-View Constraints from Videos

  New version: http://arxiv.org/abs/2012.04641 Video: http://youtube.com/watch?v=R1cXg0vpwe4

  Youtube: https://www.youtube.com/watch?v=R1cXg0vpwe4&t=4s&ab_channel=MatthiasNiessner

  reference: https://twitter.com/VittoFerrariCV/status/1486284329743527940?s=20

• ShapeFormer

  AK

  ShapeFormer: Transformer-based Shape Completion via Sparse Representation

  abs: https://arxiv.org/abs/2201.10326

  project page: https://shapeformer.github.io

  reference: https://twitter.com/ak92501/status/1486162812057628675?s=20

• Denoising diffusion GAN

  Arash Vahdat

  Denoising diffusion GAN has been accepted to #ICLR2022 as a spotlight paper

  w/ the amazing Zhisheng Xiao & @karsten_kreis Public address loudspeaker I will present this work on Friday at @ml_collective reading group organized by @savvyRL. Sign up if you're interested: https://mlcollective.org/dlct/

  reference: https://twitter.com/ArashVahdat/status/1488540465125216256?s=20&t=PHd11DgNB6ZHNdVptwULcA

• Understanding over-squashing and bottlenecks on graphs via curvature

  Jake Topping

  Thrilled that our paper "Understanding over-squashing and bottlenecks on graphs via curvature" has been accepted for an oral presentation at #ICLR2022! Big thank you to the team @Francesco_dgv @b_p_chamberlain @epomqo @mmbronstein (https://arxiv.org/abs/2111.14522)

  reference: https://twitter.com/jctopping/status/1488516500826767362?s=20&t=PHd11DgNB6ZHNdVptwULcA