I am Christian Rupprecht, Associate Professor in Computer Science at the University of Oxford and part of VGG. I am interested in unsupervised scene understanding in {2,3,4}D from images and videos. I was a PostDoc with Andrea Vedaldi and an intern with Chris Pal at the Montreal Institute For Learning Algorithms. I obtained my PhD from the Technical University of Munich, Germany supervised by Nassir Navab and Gregory D. Hager.
News
- Four papers accepted at CVPR’24: SplatterImage, VGGSfM, 3DFauna, Cache Me if You Can
- Two papers accepted at 3DV’24: Farm3D and S4C (spotlight)
- Three papers accepted at ICCV’23: RedCircle (oral), Viewset Diffusion, Pose Diffusion,
- Six papers accepted at CVPR’23: BehindTheScenes, ContinualObjectDetection, DynamicStereo, MagicPony, PointCloudDiffusion, RealFusion
- One paper on contrastive learning on long-tailed data accepted at ICLR’23
- I will be an area chair for CVPR’23
- One paper accepted at NeurIPS’22 on unsupervised video segmentation
- One paper accepted at ECCV’22 on multi-modal video understanding
- I was an invited speaker at the Rank Symposium on Neural Rendering
- I gave a lecture on “Unsupervised Learning in Computer Vision” at the BMVA summer school
- Two papers accepted at CVPR’22 on spectral clustering (oral) and on weakly supervised derendering
- Our BMVC’21 paper recieved the Best Student Paper award!
- Videos for our Unsupervised 3D Learning in the Wild workshop at ICCV21 are now online
- One paper accepted at NeurIPS 2021 on unsupervised object part segmentation
- Two papers accepted at NeurIPS 2021 Datasets & benchmarks: an ImageNet replacement without humans for pretraining and ClevrTex a synthetic multi object dataset with realistic textures
- One paper accepted at BMVC 2021 (oral) on layman supervised classification using external knowledge
- I am now a Departmental Lecturer at VGG Oxford
- One paper accepted at CVPR 2021 on finding critical paths in CNNs
- One paper on video clustering accepted at NeurIPS 2020 (code)
- Our work on unsupervised 3D reconstruction has received the Best Paper Award at CVPR 2020
- Two papers accepted at CVPR 2020: on unsupervised 3D from single image (oral, blog post, code) and semantic scene manipulation (code, project page)
- Three papers accepted at ICLR 2020: on self-labelling (spotlight, blog post, code), learning from one image and RL visualization
- Our team won 2nd place at the traffic4cast challenge at NeurIPS
- Two papers accepted at ICCV 2019: on Unsupervised Captioning and 6DoF Pose Estimation.
- One paper accepted in Nature Scientific Reports