Philipp Krähenbühl

Department of Computer Science
University of Texas at Austin
2317 Speedway
Austin, TX 78712-1757

email: philkr (at) cs.utexas.edu
CV DBLP Google Scholar github

Research

I am an Associate Professor in the Department of Computer Science at the University of Texas at Austin and Researcher at Apple. I received my PhD in 2014 from the CS Department at Stanford University and then spent two wonderful years as a PostDoc at UC Berkeley.

My research interests lie in Computer Vision, Machine Learning, and Language Models. I’m particularly interested in multimodal foundation models, interactive agents, and reinforcement learning post-training of large models.

Publications

2026
	Mask-Aware Policy Gradients for Diffusion Language Models Haran Raajesh, Kulin Shah, Adam Klivans, Philipp Krähenbühl COLM 2026 arxiv
	Spherical Leech Quantization for Visual Tokenization and Generation Yue Zhao, Hanwen Jiang, Zhenlin Xu, Chutong Yang, Ehsan Adeli, Philipp Krähenbühl CVPR 2026 arxiv project
	Latent Chain-of-Thought World Modeling for End-to-End Driving Shuhan Tan, Kashyap Chitta, Yuxiao Chen, Ran Tian, Yurong You, Yan Wang, Wenjie Luo, Yulong Cao, Philipp Krähenbühl, Marco Pavone, Boris Ivanovic CVPR 2026 arxiv
	Entropy-Preserving Reinforcement Learning Aleksei Petrenko, Ben Lipkin, Kevin Chen, Erik Wijmans, Marco Cusumano-Towner, Raja Giryes, Philipp Krähenbühl ICLR 2026 arxiv
	Triangle Multiplication Is All You Need For Biomolecular Structure Representations Jeffrey Ouyang-Zhang, Pranav Murugan, Daniel J. Diaz, Gianluca Scarpellini, Richard Strong Bowen, Nate Gruver, Adam Klivans, Philipp Krähenbühl, Aleksandra Faust, Maruan Al-Shedivat ICLR 2026 code arxiv
2025
	PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding Jang Hyun Cho, Andrea Madotto, Effrosyni Mavroudi, Triantafyllos Afouras, Tushar Nagarajan, Muhammad Maaz, Yale Song, Tengyu Ma, Shuming Hu, Suyog Jain, Miguel Martin, Huiyu Wang, Hanoona Rasheed, Peize Sun, Po-Yao Huang, Daniel Bolya, Nikhila Ravi, Shashank Jain, Tammy Stark, Shane Moon, Babak Damavandi, Vivian Lee, Andrew Westbury, Salman Khan, Philipp Krähenbühl, Piotr Dollár, Lorenzo Torresani, Kristen Grauman, Christoph Feichtenhofer NeurIPS 2025 code arxiv
	Long-term Traffic Simulation with Interleaved Autoregressive Motion and Scenario Generation Xiuyu Yang, Shuhan Tan, Philipp Krähenbühl ICCV 2025 code arxiv project
	Robust Autonomy Emerges from Self-Play Marco Cusumano-Towner, David Hafner, Alex Hertzberg, Brody Huval, Aleksei Petrenko, Eugene Vinitsky, Erik Wijmans, Taylor Killian, Stuart Bowers, Ozan Sener, Philipp Krähenbühl, Vladlen Koltun ICML 2025 arxiv
	Cut Your Losses in Large-Vocabulary Language Models Erik Wijmans, Brody Huval, Alexander Hertzberg, Vladlen Koltun, Philipp Krähenbühl ICLR 2025 code arxiv
	Distilling Structural Representations into Protein Sequence Models Jeffrey Ouyang-Zhang, Chengyue Gong, Yue Zhao, Philipp Krähenbühl, Adam R Klivans, Daniel J Diaz ICLR 2025 code arxiv
	Does Spatial Cognition Emerge in Frontier Models? Santhosh Kumar Ramakrishnan, Erik Wijmans, Philipp Kraehenbuehl, Vladlen Koltun ICLR 2025 code arxiv
	Image and Video Tokenization with Binary Spherical Quantization Yue Zhao, Yuanjun Xiong, Philipp Krähenbühl ICLR 2025 code arxiv
	Language-Image Models with 3D Understanding Jang Hyun Cho, Boris Ivanovic, Yulong Cao, Edward Schmerling, Yue Wang, Xinshuo Weng, Boyi Li, Yurong You, Philipp Krähenbühl, Yan Wang, Marco Pavone ICLR 2025 arxiv project
2024
	Distilling Vision-Language Models on Millions of Videos Yue Zhao, Long Zhao, Xingyi Zhou, Jialin Wu, Chun-Te Chu, Hui Miao, Florian Schroff, Hartwig Adam, Ting Liu, Boqing Gong, Philipp Krähenbühl, Liangzhe Yuan CVPR 2024 arxiv
	Promptable Closed-loop Traffic Simulation Shuhan Tan, Boris Ivanovic, Yuxiao Chen, Boyi Li, Xinshuo Weng, Yulong Cao, Philipp Krähenbühl, Marco Pavone CoRL 2024 arxiv
2023
	PartDistillation: Learning Parts from Instance Segmentation Jang Hyun Cho, Philipp Krähenbühl, Vignesh Ramanathan CVPR 2023 code pdf
	Learning Video Representations from Large Language Models Yue Zhao, Ishan Misra, Philipp Krähenbühl, Rohit Girdhar CVPR 2023 code arxiv
	Language Conditioned Traffic Generation Shuhan Tan, Boris Ivanovic, Xinshuo Weng, Marco Pavone, Philipp Kraehenbuehl CoRL 2023 code arxiv
	Language-conditioned Detection Transformer Jang Hyun Cho, Philipp Krähenbühl CVPR 2023 code arxiv
	Predicting a Protein's Stability under a Million Mutations Jeffrey Ouyang-Zhang, Daniel J. Diaz, Adam R. Klivans, Philipp Krähenbühl NeurIPS 2023 code arxiv
2022
	Real-Time Online Video Detection with Temporal Smoothing Transformers Yue Zhao, Philipp Krähenbühl ECCV 2022 code arxiv
	Long-tail detection with effective class-margins Jang Hyun Cho, Philipp Krähenbühl ECCV 2022 code arxiv
	Detecting twenty-thousand classes using image-level supervision Xingyi Zhou, Rohit Girdhar, Armand Joulin, Philipp Krähenbühl, Ishan Misra ECCV 2022 code arxiv
	Learning from All Vehicles Dian Chen, Philipp Krähenbühl CVPR 2022 code arxiv
	Cross-view Transformers for real-time Map-view Semantic Segmentation Brady Zhou, Philipp Krähenbühl CVPR 2022 code arxiv
	Global Tracking Transformers Xingyi Zhou, Tianwei Yin, Vladlen Koltun, Philipp Krähenbühl CVPR 2022 code arxiv
	Simple multi-dataset detection Xingyi Zhou, Vladlen Koltun, Philipp Krähenbühl CVPR 2022 code arxiv
2021
	Multimodal Virtual Point 3D Detection Tianwei Yin, Xingyi Zhou, Philipp Krähenbühl NeurIPS 2021 code arxiv
	Learning to drive from a world on rails Dian Chen, Vladlen Koltun, Philipp Krähenbühl ICCV 2021 arxiv
	Towards Long-Form Video Understanding Chao-Yuan Wu, Philipp Krähenbühl CVPR 2021 arxiv
	Center-based 3d object detection and tracking Tianwei Yin, Xingyi Zhou, Philipp Krähenbühl CVPR 2021 code arxiv
	Memory Optimization for Deep Networks Aashaka Shah, Chao-Yuan Wu, Jayashree Mohan, Vijay Chidambaram, Philipp Krähenbühl ICLR 2021 code arxiv
	Probabilistic two-stage detection Xingyi Zhou, Vladlen Koltun, Philipp Krähenbühl arXiv 2021 arxiv
2020
	Domain Adaptation Through Task Distillation Brady Zhou, Nimit Kalra, Philipp Krähenbühl ECCV 2020 code arxiv
	Tracking Objects as Points Xingyi Zhou, Vladlen Koltun, Philipp Krähenbühl ECCV 2020 code arxiv
	A Multigrid Method for Efficiently Training Video Models Chao-Yuan Wu, Ross Girshick, Kaiming He, Christoph Feichtenhofer, Philipp Krähenbühl CVPR 2020 code arxiv
2019
	Learning by Cheating Dian Chen, Brady Zhou, Vladlen Koltun, Philipp Krähenbühl CORL 2019 code arxiv
	Monocular plan view networks for autonomous driving Dequan Wang, Coline Devin, Qi-Zhi Cai, Philipp Krähenbühl, Trevor Darrell IROS 2019 arxiv
	Objects as points Xingyi Zhou, Dequan Wang, Philipp Krähenbühl arXiv preprint arXiv:1904.07850 2019 code arxiv
	Long-Term Feature Banks for Detailed Video Understanding Chao-Yuan Wu, Christoph Feichtenhofer, Haoqi Fan, Kaiming He, Philipp Krähenbühl, Ross Girshick CVPR 2019 code arxiv supplement
	Bottom-up Object Detection by Grouping Extreme and Center Points Xingyi Zhou, Jiacheng Zhuo, Philipp Krähenbühl CVPR 2019 code arxiv supplement
	Joint Monocular 3D Vehicle Detection and Tracking Hou-Ning Hu, Qi-Zhi Cai, Dequan Wang, Ji Lin, Min Sun, Philipp Krähenbühl, Trevor Darrell, Fisher Yu ICCV 2019 code arxiv
	Does Computer Vision Matter for Action? Brady Zhou, Philipp Krähenbühl, Vladlen Koltun Science Robotics 2019 code arxiv
	Don't let your Discriminator be fooled Brady Zhou, Philipp Krähenbühl ICLR 2019 pdf
2018
	Video Compression through Image Interpolation Chao-Yuan Wu, Nayan Singhal, Philipp Krähenbühl ECCV 2018 code arxiv
	Domain transfer through deep activation matching Haoshuo Huang, Qixing Huang, Philipp Krähenbühl ECCV 2018 pdf project
	Compressed Video Action Recognition Chao-Yuan Wu,Manzil Zaheer,Hexiang Hu,R. Manmatha,Alexander J. Smola, Philipp Krähenbühl CVPR 2018 code arxiv project
	Free Supervision from Video Games Philipp Krähenbühl CVPR 2018 code pdf project
	Assessing Generalization in Deep Reinforcement Learning Charles Packer, Katelyn Gao, Jernej Kos, Philipp Krähenbühl, Vladlen Koltun, Dawn Song arXiv 2018 arxiv
2017
	Sampling Matters in Deep Embedding Learning Chao-Yuan Wu, R. Manmatha, Alexander J. Smola, Philipp Krähenbühl ICCV 2017 code arxiv project
	Adversarial Feature Learning Jeff Donahue, Philipp Krähenbühl, Trevor Darrell ICLR 2017 code arxiv
2016
	Generative Visual Manipulation on the Natural Image Manifold Jun-Yan Zhu, Philipp Krähenbühl, Eli Shechtman, Alexei A. Efros ECCV 2016 code arxiv project
	Context Encoders: Feature Learning by Inpainting Deepak Pathak, Philipp Krähenbühl, Jeff Donahue, Trevor Darrell, Alyosha Efros CVPR 2016 code arxiv project
	Learning Dense Correspondence via 3D-guided Cycle Consistency Tinghui Zhou, Philipp Krähenbühl, Mathieu Aubry, Qixing Huang, Alyosha Efros CVPR 2016 arxiv project
	Data-dependent initializations of convolutional neural networks Philipp Krähenbühl, Carl Doersch, Jeff Donahue, Trevor Darrell ICLR 2016 code arxiv
2015
	Learning a Discriminative Model for the Perception of Realism in Composite Images Jun-Yan Zhu, Philipp Krähenbühl, Eli Shechtman, Alyosha Efros ICCV 2015 code arxiv
	Learning Data-driven Reflectance Priors for Intrinsic Image Decomposition Tinghui Zhou, Philipp Krähenbühl, Alyosha Efros ICCV 2015 code arxiv
	Constrained Convolutional Neural Networks for Weakly Supervised Segmentation Deepak Pathak, Philipp Krähenbühl, Trevor Darrell ICCV 2015 code arxiv supplement
	Learning to propose objects Philipp Krähenbühl, Vladlen Koltun CVPR 2015 code pdf
2014
	Geodesic Object Proposals Philipp Krähenbühl, Vladlen Koltun ECCV 2014 code pdf
2013
	Parameter Learning and Convergent Inference for Dense Random Fields Philipp Krähenbühl, Vladlen Koltun ICML 2013 code pdf project
2012
	Efficient Nonlocal regularization for Optical Flow Philipp Krähenbühl, Vladlen Koltun ECCV 2012 pdf
	Saliency Filters: Contrast Based Filtering for Salient Region Detection Federico Perazzi, Philipp Krähenbühl, Yael Pritch, Alexander Hornung CVPR 2012 code pdf project
2011
	Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials Philipp Krähenbühl, Vladlen Koltun NIPS 2011 code arxiv project
2010
	Gesture Controllers Sergey Levine, Philipp Krähenbühl, Sebastian Thrun, Vladlen Koltun SIGGRAPH 2010 pdf
2009
	A system for retargeting of streaming video Philipp Krähenbühl, Manuel Lang, Alexander Hornung, Markus Gross SIGGRAPH Asia 2009 pdf

Research group

PhD Students:

Past PhD students:

Jeffrey Ouyang-Zhang (2026, next: Postdoc at MIT/Biohub)
Brady Zhou (2026)
Shuhan Tan (2026, next: NVIDIA)
Jang Hyun (Vincent) Cho (2025, next: Meta, now: Thinking Machines)
Yue Zhao (2025, next: Postdoc at Stanford)
Dian Chen (2023, next: Waabi)
Xingyi Zhou (2022, next: Google, now: AMI)
Chao-Yuan Wu (2021, next: Facebook, now: World Labs)

Undergraduates and MS:

Past undergraduates and MS:

Tianwei Yin (next: MIT)
Scott Cao (next: Facebook)
David Wang (next: Amazon)
Chia-Wen Cheng (next: Facebook)
Mina Lee (next: Google)
Kamil Ali (next: Stanford)
Brady Zhou (next: Intel, then UT)
Nayan Singhal (next: Facebook AML)
Shaayaan Sayed (next: some hedgefund)
Nimit Kalra (next: finance)

Teaching

Joining my research group

UT CS or ECE students: I’d recomment you to take my graduate deep learning class (CS395T), and start working with me through that class.

Prospective students: Please read about our graduate admissions process and state your interested in my research group in your statement of purpose. Please do *not- contact me directly. The statistics are not in your favor either. We have not yet admitted a single student to UTCS who contacted me directly.

About my last name

I’m well aware that my last name is not the easiest one to write or cite (and I saw it butchered a bunch of times over the years). So to make things easier just pick your document type below and copy the string:

Regular text

Krähenbühl

Latex & Bibtex

Kr\"ahenb\"uhl

HTML

Kr&auml;henb&uuml;hl

If all the above fail, just use Kraehenbuehl.