This video summarizes the #ECCV2022 papers we are going to present in the next three days. If you are interested in egocentric human body estimation, scalable human avatar creation, and human motion and interaction synthesis, don't miss it. European Conference on Computer Vision #ECCV2026 youtube.com/watch?v=WNF9yt…

We are presenting our paper #EgoBody Dataset at #ECCV2022! Drop by our poster (#94) in Wednesday morning session (hall B) 11:00-13:30 for a chat!

PlanT: Explainable Planning Transformers via Object-Level Representations abs: arxiv.org/abs/2210.14222 project page: katrinrenz.de/plant/

📢Check out our KeypointNeRF at #ECCV2022 in hall B today from 11:00-13:30!! 🏖️🏖️ Project: markomih.github.io/KeypointNeRF/ Overview: youtu.be/RMs1S5k9vrk We present a high-res and generalizable NeRF for human face and body without requiring 3DMM or SMPL. 🧵⬇️ (1/N)

NeRFPlayer: A Streamable Dynamic Scene Representation with Decomposed Neural Radiance Fields abs: arxiv.org/abs/2210.15947 project page: lsongx.github.io/projects/nerfp…

Is it possible to construct the real world virtually? #ETHZurich researchers have developed a new approach to this question. The Sony Research Award Program supports the project by Songyou Peng and team. #VirtualWorlds #VR youtube.com/watch?v=hsGJTD…

Are RGB inputs good enough for open-world generalization of object detection? Excited to share our #ICLR2023 paper GOOD -- we show that geometric cues can significantly boost the performance for the task! arxiv.org/abs/2212.11720 (Joint work with @AutoVisionGroup Dan Zhang )

In less than an hour I am going to present our paper MERF at #SIGGRAPH2023 in Petree Hall D. MERF allows you to interactively explore large scenes on a laptop in the browser. Check out our web demo: creiser.github.io/merf/ By the way we have also released the entire code now!

Introducing SMERF: a streamable, memory-efficient method for real-time exploration of large, multi-room scenes on everyday devices. Our method brings the realism of Zip-NeRF to your phone or laptop! Project page: smerf-3d.github.io ArXiv: arxiv.org/abs/2312.07541 (1/n)

Ever wondered why RL is used in Gemini and ChatGPT even though LLMs are considered (self-)supervised terrain? The key is that RL can optimize non-differentiable objectives, like human feedback! We revisit RL from this alternative angle in our new tutorial: arxiv.org/abs/2312.08365

I’m very happy to share another recent work: Diffusion Noise Optimization (DNO). We demonstrate how to optimize diffusion latent noise using criterion functions defined in motion space, which serves as universal motion priors for a wide range of motion-related tasks.

Happy to share our #ICLR2024 paper with Yuanwen Yue, Sabarinath Mahadevan,Jonas Schult,Francis Engelmann, Bastian Leibe, Konrad Schindler! AGILE3D provides multiple 3D instance masks simultaneously with a few user clicks. 📖 arxiv.org/abs/2306.00977 ⌨️ ywyue.github.io/AGILE3D/

🚀Excited to share our recent work on test-time scaling for feed-forward Gaussian splatting: we learn a recurrent model ReSplat that is able to iteratively improve the reconstruction quality in a feed-forward manner! haofeixu.github.io/resplat/