SOFTWARE & DATASETS


2022

Software for Learning Goal-Conditioned Policies Offline

[CoRL 2022 paper][Project page][Code]
Software for Bird's-Eye-View Segmentation

[CoRL 2022 paper][Project page][Code]
VisSpeech Dataset

[INTERSPEECH 2022 paper][Project page][Dataset]

2021

Evaluation of ImageNet-CoG Benchmark

[ICCV 2021 paper][Project page][Code]

2020

Software for Multimodal Transformer

[ECCV 2020 paper][Project page][Code]

2019

Software for Motion-Augmented RGB for Action Recognition

[CVPR 2019 paper][Project page][Code]
Software for end-to-end incremental learning

[ECCV 2018 paper][Code]

2018

Software for Incremental Object Detectors

[ICCV 2017 paper][Code]
Charades-Ego Dataset

[CVPR 2018 paper][Project page][Dataset]
Software for joint modeling of first and third-person videos

[CVPR 2018 paper][Code]

2017

Software for learning video object segmentation

[ICCV 2017 paper][Journal version][Project page][Code]
Software for learning motion patterns

[CVPR 2017 paper][Project page][Code]
Software for online object tracking

[ICCV 2015 paper][Project page][Code]

2016

Software for weakly-supervised semantic segmentation

[ECCV 2016 paper][Project page][Code]

2015

Software for pose estimation and segmentation of multiple people

[TPAMI 2015 paper][Project page][Code]

2014

Software for estimating human pose in videos

[CVPR 2014 paper][Project page][Code]
Poses in the Wild Dataset
This dataset has 30 video sequences generated from three Hollywood movies. Each sequence has approximately 30 frames and is annotated for human upper-body keypoints.

[CVPR 2014 paper][Project page][Dataset]

2013

Inria 3DMovie Dataset
This dataset contains all the stereo pairs and their annotations used in our ICCV 2013 paper. Most of this data was extracted from the "StreetDance 3D" [Giwa and Pasquini, 2010] and "Pina" [Wenders, 2011] stereo movies.

[ICCV 2013 paper][Project page][Dataset]
Software for learning graphs

[ICCV 2013 paper][Code]
WILLOW-ObjectClass dataset for evaluating graph matching
A dataset containing object class images and their part annotations.

[ICCV 2013 paper][Dataset]
IIIT-STR, Sports-10K, TV Series-1M scene text retrieval datasets
The IIIT STR dataset is harvested from Google and Flickr image search. Sports-10K dataset is from sports video clips, containing many advertisement signboards, and the TV Series-1M is from four popular TV series: Friends, Buffy, Mr. Bean, and Open All Hours.

[ICCV 2013 paper][Project page][Image & Video datasets]

2012

Software for solving detection and segmentation problems jointly

[ECCV 2010 paper][Code]
IIIT 5K-Word dataset
This dataset is harvested from Google image search. It contains 5000 cropped word images from Scene Texts and born-digital images.

[BMVC 2012 paper][Project page][Dataset]
SVT-CHAR: Annotated character dataset
This dataset contains character level bounding boxes and ground truth annotation of SVT-WORD dataset.

[CVPR 2012 paper][Project page][Dataset]

2011

Alpha-expansion Beta-shrink Moves Software for MRFs

[UAI 2011 paper][Code]
INRIA-Video dataset
A dataset containing short video clips and annotation of a subset of frames for evaluating video segmentation.

[CVPR 2011 paper][Project page][Dataset]

2010

Dataset for Vision labelling problems
Data used in our TPAMI 2010 paper for evaluating (colour/object) segmentation, stereo estimation problems defined on MRF/CRF.

[TPAMI 2010 paper][Data]

2009

Efficient Solvers for Multi-label MRFs

[TPAMI 2010 paper][Code]