Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Josef Sivic

Josef Sivic

Learning to Answer Visual Questions from Web Videos


May 11, 2022
Antoine Yang, Antoine Miech, Josef Sivic, Ivan Laptev, Cordelia Schmid

Figure 1 for Learning to Answer Visual Questions from Web Videos
Figure 2 for Learning to Answer Visual Questions from Web Videos
Figure 3 for Learning to Answer Visual Questions from Web Videos
Figure 4 for Learning to Answer Visual Questions from Web Videos
* Accepted at the TPAMI Special Issue on the Best Papers of ICCV 2021. Journal extension of the conference paper arXiv:2012.00451. 16 pages, 13 figures 
  Access Paper or Ask Questions

Focal Length and Object Pose Estimation via Render and Compare


Apr 11, 2022
Georgy Ponimatkin, Yann Labbé, Bryan Russell, Mathieu Aubry, Josef Sivic

Figure 1 for Focal Length and Object Pose Estimation via Render and Compare
Figure 2 for Focal Length and Object Pose Estimation via Render and Compare
Figure 3 for Focal Length and Object Pose Estimation via Render and Compare
Figure 4 for Focal Length and Object Pose Estimation via Render and Compare
* Accepted to CVPR2022. Code available at http://github.com/ponimatkin/focalpose 
  Access Paper or Ask Questions

TubeDETR: Spatio-Temporal Video Grounding with Transformers


Mar 30, 2022
Antoine Yang, Antoine Miech, Josef Sivic, Ivan Laptev, Cordelia Schmid

Figure 1 for TubeDETR: Spatio-Temporal Video Grounding with Transformers
Figure 2 for TubeDETR: Spatio-Temporal Video Grounding with Transformers
Figure 3 for TubeDETR: Spatio-Temporal Video Grounding with Transformers
Figure 4 for TubeDETR: Spatio-Temporal Video Grounding with Transformers
* Accepted at CVPR 2022 (Oral); 17 pages; 8 figures 
  Access Paper or Ask Questions

Look for the Change: Learning Object States and State-Modifying Actions from Untrimmed Web Videos


Mar 22, 2022
Tomáš Souček, Jean-Baptiste Alayrac, Antoine Miech, Ivan Laptev, Josef Sivic

Figure 1 for Look for the Change: Learning Object States and State-Modifying Actions from Untrimmed Web Videos
Figure 2 for Look for the Change: Learning Object States and State-Modifying Actions from Untrimmed Web Videos
Figure 3 for Look for the Change: Learning Object States and State-Modifying Actions from Untrimmed Web Videos
Figure 4 for Look for the Change: Learning Object States and State-Modifying Actions from Untrimmed Web Videos
* To be published in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2022 
  Access Paper or Ask Questions

Drive&Segment: Unsupervised Semantic Segmentation of Urban Scenes via Cross-modal Distillation


Mar 21, 2022
Antonin Vobecky, David Hurych, Oriane Siméoni, Spyros Gidaris, Andrei Bursuc, Patrick Pérez, Josef Sivic

Figure 1 for Drive&Segment: Unsupervised Semantic Segmentation of Urban Scenes via Cross-modal Distillation
Figure 2 for Drive&Segment: Unsupervised Semantic Segmentation of Urban Scenes via Cross-modal Distillation
Figure 3 for Drive&Segment: Unsupervised Semantic Segmentation of Urban Scenes via Cross-modal Distillation
Figure 4 for Drive&Segment: Unsupervised Semantic Segmentation of Urban Scenes via Cross-modal Distillation
* See project webpage https://vobecant.github.io/DriveAndSegment/ for the code and more 
  Access Paper or Ask Questions

Learning to Manipulate Tools by Aligning Simulation to Video Demonstration


Nov 04, 2021
Kateryna Zorina, Justin Carpentier, Josef Sivic, VladimĂ­r PetrĂ­k

Figure 1 for Learning to Manipulate Tools by Aligning Simulation to Video Demonstration
Figure 2 for Learning to Manipulate Tools by Aligning Simulation to Video Demonstration
Figure 3 for Learning to Manipulate Tools by Aligning Simulation to Video Demonstration
Figure 4 for Learning to Manipulate Tools by Aligning Simulation to Video Demonstration
* Accepted to IEEE Robotics and Automation Letters (RA-L) 
  Access Paper or Ask Questions

Estimating 3D Motion and Forces of Human-Object Interactions from Internet Videos


Nov 02, 2021
Zongmian Li, Jiri Sedlar, Justin Carpentier, Ivan Laptev, Nicolas Mansard, Josef Sivic

Figure 1 for Estimating 3D Motion and Forces of Human-Object Interactions from Internet Videos
Figure 2 for Estimating 3D Motion and Forces of Human-Object Interactions from Internet Videos
Figure 3 for Estimating 3D Motion and Forces of Human-Object Interactions from Internet Videos
Figure 4 for Estimating 3D Motion and Forces of Human-Object Interactions from Internet Videos
* arXiv admin note: substantial text overlap with arXiv:1904.02683 
  Access Paper or Ask Questions

Weakly Supervised Human-Object Interaction Detection in Video via Contrastive Spatiotemporal Regions


Oct 07, 2021
Shuang Li, Yilun Du, Antonio Torralba, Josef Sivic, Bryan Russell

Figure 1 for Weakly Supervised Human-Object Interaction Detection in Video via Contrastive Spatiotemporal Regions
Figure 2 for Weakly Supervised Human-Object Interaction Detection in Video via Contrastive Spatiotemporal Regions
Figure 3 for Weakly Supervised Human-Object Interaction Detection in Video via Contrastive Spatiotemporal Regions
Figure 4 for Weakly Supervised Human-Object Interaction Detection in Video via Contrastive Spatiotemporal Regions

  Access Paper or Ask Questions

Reconstructing and grounding narrated instructional videos in 3D


Sep 10, 2021
Dimitri Zhukov, Ignacio Rocco, Ivan Laptev, Josef Sivic, Johannes L. Schönberger, Bugra Tekin, Marc Pollefeys

Figure 1 for Reconstructing and grounding narrated instructional videos in 3D
Figure 2 for Reconstructing and grounding narrated instructional videos in 3D
Figure 3 for Reconstructing and grounding narrated instructional videos in 3D
Figure 4 for Reconstructing and grounding narrated instructional videos in 3D

  Access Paper or Ask Questions

Single-view robot pose and joint angle estimation via render & compare


Apr 19, 2021
Yann Labbé, Justin Carpentier, Mathieu Aubry, Josef Sivic

Figure 1 for Single-view robot pose and joint angle estimation via render & compare
Figure 2 for Single-view robot pose and joint angle estimation via render & compare
Figure 3 for Single-view robot pose and joint angle estimation via render & compare
Figure 4 for Single-view robot pose and joint angle estimation via render & compare
* Accepted at CVPR 2021 (Oral) 
  Access Paper or Ask Questions

Visualizing computation in large-scale cellular automata


Apr 01, 2021
Hugo Cisneros, Josef Sivic, Tomas Mikolov

Figure 1 for Visualizing computation in large-scale cellular automata
Figure 2 for Visualizing computation in large-scale cellular automata
Figure 3 for Visualizing computation in large-scale cellular automata
Figure 4 for Visualizing computation in large-scale cellular automata
* Artificial Life Conference Proceedings 2020 (pp. 239-247). MIT Press 
  Access Paper or Ask Questions

Thinking Fast and Slow: Efficient Text-to-Visual Retrieval with Transformers


Mar 30, 2021
Antoine Miech, Jean-Baptiste Alayrac, Ivan Laptev, Josef Sivic, Andrew Zisserman

Figure 1 for Thinking Fast and Slow: Efficient Text-to-Visual Retrieval with Transformers
Figure 2 for Thinking Fast and Slow: Efficient Text-to-Visual Retrieval with Transformers
Figure 3 for Thinking Fast and Slow: Efficient Text-to-Visual Retrieval with Transformers
Figure 4 for Thinking Fast and Slow: Efficient Text-to-Visual Retrieval with Transformers
* Accepted to CVPR 2021 
  Access Paper or Ask Questions

Just Ask: Learning to Answer Questions from Millions of Narrated Videos


Dec 01, 2020
Antoine Yang, Antoine Miech, Josef Sivic, Ivan Laptev, Cordelia Schmid

Figure 1 for Just Ask: Learning to Answer Questions from Millions of Narrated Videos
Figure 2 for Just Ask: Learning to Answer Questions from Millions of Narrated Videos
Figure 3 for Just Ask: Learning to Answer Questions from Millions of Narrated Videos
Figure 4 for Just Ask: Learning to Answer Questions from Millions of Narrated Videos
* 19 pages; 12 figures 
  Access Paper or Ask Questions

Learning Object Manipulation Skills via Approximate State Estimation from Real Videos


Nov 13, 2020
VladimĂ­r PetrĂ­k, Makarand Tapaswi, Ivan Laptev, Josef Sivic

Figure 1 for Learning Object Manipulation Skills via Approximate State Estimation from Real Videos
Figure 2 for Learning Object Manipulation Skills via Approximate State Estimation from Real Videos
Figure 3 for Learning Object Manipulation Skills via Approximate State Estimation from Real Videos
Figure 4 for Learning Object Manipulation Skills via Approximate State Estimation from Real Videos
* CoRL 2020, code at https://github.com/makarandtapaswi/Real2Sim_CoRL2020, project page at https://data.ciirc.cvut.cz/public/projects/2020Real2Sim/ 
  Access Paper or Ask Questions

CosyPose: Consistent multi-view multi-object 6D pose estimation


Aug 19, 2020
Yann Labbé, Justin Carpentier, Mathieu Aubry, Josef Sivic

Figure 1 for CosyPose: Consistent multi-view multi-object 6D pose estimation
Figure 2 for CosyPose: Consistent multi-view multi-object 6D pose estimation
Figure 3 for CosyPose: Consistent multi-view multi-object 6D pose estimation
Figure 4 for CosyPose: Consistent multi-view multi-object 6D pose estimation
* ECCV 2020 
  Access Paper or Ask Questions

RareAct: A video dataset of unusual interactions


Aug 03, 2020
Antoine Miech, Jean-Baptiste Alayrac, Ivan Laptev, Josef Sivic, Andrew Zisserman

Figure 1 for RareAct: A video dataset of unusual interactions
Figure 2 for RareAct: A video dataset of unusual interactions
Figure 3 for RareAct: A video dataset of unusual interactions
Figure 4 for RareAct: A video dataset of unusual interactions

  Access Paper or Ask Questions

Occlusion resistant learning of intuitive physics from videos


Apr 30, 2020
Ronan Riochet, Josef Sivic, Ivan Laptev, Emmanuel Dupoux

Figure 1 for Occlusion resistant learning of intuitive physics from videos
Figure 2 for Occlusion resistant learning of intuitive physics from videos
Figure 3 for Occlusion resistant learning of intuitive physics from videos
Figure 4 for Occlusion resistant learning of intuitive physics from videos

  Access Paper or Ask Questions

Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions


Apr 22, 2020
Ignacio Rocco, Relja Arandjelović, Josef Sivic

Figure 1 for Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions
Figure 2 for Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions
Figure 3 for Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions
Figure 4 for Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions

  Access Paper or Ask Questions

End-to-End Learning of Visual Representations from Uncurated Instructional Videos


Jan 17, 2020
Antoine Miech, Jean-Baptiste Alayrac, Lucas Smaira, Ivan Laptev, Josef Sivic, Andrew Zisserman

Figure 1 for End-to-End Learning of Visual Representations from Uncurated Instructional Videos
Figure 2 for End-to-End Learning of Visual Representations from Uncurated Instructional Videos
Figure 3 for End-to-End Learning of Visual Representations from Uncurated Instructional Videos
Figure 4 for End-to-End Learning of Visual Representations from Uncurated Instructional Videos

  Access Paper or Ask Questions

Evolving Structures in Complex Systems


Nov 04, 2019
Hugo Cisneros, Josef Sivic, Tomas Mikolov

Figure 1 for Evolving Structures in Complex Systems
Figure 2 for Evolving Structures in Complex Systems
Figure 3 for Evolving Structures in Complex Systems
Figure 4 for Evolving Structures in Complex Systems
* Proceedings of the 2019 IEEE Symposium Series on Computational Intelligence 
 * IEEE Symposium Series on Computational Intelligence 2019 (IEEE SSCI 2019) 
  Access Paper or Ask Questions

Is This The Right Place? Geometric-Semantic Pose Verification for Indoor Visual Localization


Sep 02, 2019
Hajime Taira, Ignacio Rocco, Jiri Sedlar, Masatoshi Okutomi, Josef Sivic, Tomas Pajdla, Torsten Sattler, Akihiko Torii

Figure 1 for Is This The Right Place? Geometric-Semantic Pose Verification for Indoor Visual Localization
Figure 2 for Is This The Right Place? Geometric-Semantic Pose Verification for Indoor Visual Localization
Figure 3 for Is This The Right Place? Geometric-Semantic Pose Verification for Indoor Visual Localization
Figure 4 for Is This The Right Place? Geometric-Semantic Pose Verification for Indoor Visual Localization

  Access Paper or Ask Questions

Combining learned skills and reinforcement learning for robotic manipulations


Aug 02, 2019
Robin Strudel, Alexander Pashevich, Igor Kalevatykh, Ivan Laptev, Josef Sivic, Cordelia Schmid

Figure 1 for Combining learned skills and reinforcement learning for robotic manipulations
Figure 2 for Combining learned skills and reinforcement learning for robotic manipulations
Figure 3 for Combining learned skills and reinforcement learning for robotic manipulations
Figure 4 for Combining learned skills and reinforcement learning for robotic manipulations
* 11 pages 
  Access Paper or Ask Questions

HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips


Jul 31, 2019
Antoine Miech, Dimitri Zhukov, Jean-Baptiste Alayrac, Makarand Tapaswi, Ivan Laptev, Josef Sivic

Figure 1 for HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips
Figure 2 for HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips
Figure 3 for HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips
Figure 4 for HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips
* Accepted at ICCV 2019 
  Access Paper or Ask Questions

Temporal Localization of Moments in Video Collections with Natural Language


Jul 30, 2019
Victor Escorcia, Mattia Soldan, Josef Sivic, Bernard Ghanem, Bryan Russell

Figure 1 for Temporal Localization of Moments in Video Collections with Natural Language
Figure 2 for Temporal Localization of Moments in Video Collections with Natural Language
Figure 3 for Temporal Localization of Moments in Video Collections with Natural Language
Figure 4 for Temporal Localization of Moments in Video Collections with Natural Language

  Access Paper or Ask Questions

D2-Net: A Trainable CNN for Joint Detection and Description of Local Features


May 09, 2019
Mihai Dusmanu, Ignacio Rocco, Tomas Pajdla, Marc Pollefeys, Josef Sivic, Akihiko Torii, Torsten Sattler

Figure 1 for D2-Net: A Trainable CNN for Joint Detection and Description of Local Features
Figure 2 for D2-Net: A Trainable CNN for Joint Detection and Description of Local Features
Figure 3 for D2-Net: A Trainable CNN for Joint Detection and Description of Local Features
Figure 4 for D2-Net: A Trainable CNN for Joint Detection and Description of Local Features
* Accepted at CVPR 2019 
  Access Paper or Ask Questions

Monte-Carlo Tree Search for Efficient Visually Guided Rearrangement Planning


Apr 23, 2019
Sergey Zagoruyko, Yann Labbé, Igor Kalevatykh, Ivan Laptev, Justin Carpentier, Mathieu Aubry, Josef Sivic

Figure 1 for Monte-Carlo Tree Search for Efficient Visually Guided Rearrangement Planning
Figure 2 for Monte-Carlo Tree Search for Efficient Visually Guided Rearrangement Planning
Figure 3 for Monte-Carlo Tree Search for Efficient Visually Guided Rearrangement Planning
Figure 4 for Monte-Carlo Tree Search for Efficient Visually Guided Rearrangement Planning

  Access Paper or Ask Questions

Estimating 3D Motion and Forces of Person-Object Interactions from Monocular Video


Apr 04, 2019
Zongmian Li, Jiri Sedlar, Justin Carpentier, Ivan Laptev, Nicolas Mansard, Josef Sivic

Figure 1 for Estimating 3D Motion and Forces of Person-Object Interactions from Monocular Video
Figure 2 for Estimating 3D Motion and Forces of Person-Object Interactions from Monocular Video
Figure 3 for Estimating 3D Motion and Forces of Person-Object Interactions from Monocular Video
Figure 4 for Estimating 3D Motion and Forces of Person-Object Interactions from Monocular Video

  Access Paper or Ask Questions