Computer Vision research group @UniofOxford led by Andrew Zisserman, Andrea Vedaldi, João Henriques, Christian Rupprecht, and Iro Laina
Oxford, England
- Tweets 1,123
- Following 367
- Followers 14,502
- Likes 920
Pinned Tweet
We are seeking a full-time Postdoctoral Research Assistant in Computer Vision to join the Visual Geometry Group (University of Oxford) to work on 3D and Spatial AI with Professor Andrea Vedaldi. The post is funded by ERC and is fixed-term for two years with a possible extension.
3
13
54
15,028
Many Congratulations to @jianyuan_wang, @MinghaoChen23, @n_karaev, Andrea Vedaldi, Christian Rupprecht and @davnov134 for winning the Best Paper Award @CVPR for "VGGT: Visual Geometry Grounded Transformer" 🥇🎉 🙌🙌 #CVPR2025!!!!!!
17
68
488
46,283
Congratulations to @elliottszwu @chrirupp and Andrea Vedaldi on winning the Best Paper Award #CVPR2020!!!
5
39
239
Congratulations to Andrea Vedaldi and Brian Fulkerson on receiving the 2020 ACM SIGMM Test of Time Paper Award for VLFeat (vlfeat.org/)!
@sigmm
2
37
227
SOTA results on unsupervised video representation learning (beating supervised pretraining)!
Excited to announce new work "Multi-Modal Self-Supervision from Generalized Data Transformations" by @mandelapatrick_, @y_m_asano et al, in colab w/ @facebookai
arxiv.org/abs/2003.04298
2
45
141
Congratulations to @jianyuan_wang, @MinghaoChen23, @n_karaev, Andrea Vedaldi, Christian Rupprecht and @davnov134 for "VGGT: Visual Geometry Grounded Transformer" for being an Award Candidate @CVPR. #CVPR2025 🎉 🙌. cvpr.thecvf.com/virtual/2025…
2
10
140
7,119
Congratulations to @LiliMomeni, @gulvarol, @samuelalbanie, @afourast and Andrew Zisserman, who received the Best Application Paper Award @ACCV2020 #ACCV2020!
1
13
114
🚗🚗 You drive past a parked car. What does the side you didn't see look like? Let SNeS (Symmetric Neural Surfaces) show you at #ECCV2022!
📰 Blog: robots.ox.ac.uk/~vgg/blog/sn…
🚩 Project Page: robots.ox.ac.uk/~vgg/researc…
🪄 Code: github.com/eldar/snes
📝 Paper: arxiv.org/abs/2206.06340
2
23
119
Tired of reviewers obsessing over state of the art results?
A new #ICCV2019 workshop is launching to trial the preregistration model in computer vision, i.e.:
Idea -> Paper -> Reviews -> Experiments
More info at preregister.vision!
1
41
106
Andreas Geiger (@uni_tue, @MPI_IS) visited us virtually and gave an amazing talk on Learning 3D Reconstruction in Function Space, covering an incredible line of research with implicit functions!
Video recording and slides are available: piped.video/watch?v=kxKI8_Si…
#VGGVisitors
26
103
Self-Labelling via simultaneous clustering and representation learning. #ICLR20 @iclr_conf paper.
We learn meaningful labels and get SOTA without NCE. @y_m_asano, @chrirupp, Vedaldi
👉Blog: robots.ox.ac.uk/~vgg/blog/se…
👨💻Code: github.com/yukimasano/self-l…
🌐Page: robots.ox.ac.uk/~vgg/researc…
3
38
100
Tired of reviewers obsessing over state of the art results?
Submit a paper without experiments (but with a plan for experiments) to the @NeurIPSConf 2021 Pre-registration Workshop!
Deadline: 17 Sept this Friday!
preregister.science
2
16
99
Check out our new #ICLR2021 spotlight paper on video-text representation learning: arxiv.org/abs/2010.02824
We add a generative objective to the usual contrastive instance disc. loss and get SoTA on numerous datasets and tasks.
Work by @mandelapatrick_ @y_m_asano et al @iclr_conf
18
94
'Invariant Information Clustering'-new work by @xu__ji and Andrea Vedaldi on unsupervised #imageclassification and #segmentation! Beats #unsupervisedlearning state-of-the-art on 8 datasets incl. CIFAR, MNIST, COCO-Stuff & STL from ImageNet. Paper and code: arxiv.org/abs/1807.06653
31
88
Congratulations @gulvarol for winning the ELLIS PhD Award 2020 for her thesis - “Learning Human Body and Human Action Representations from Visual Data” supervised by @CordeliaSchmid and Ivan Laptev! ellis.eu/news/ellis-phd-awar… @ELLISforEurope #AcademicChatter #PhDaward
3
5
77
New #ICCV2019 paper and PyTorch library on interpretability! Understanding Deep Networks via Extremal Perturbations and Smooth Masks by @ruthiecf @mandelapatrick_ and Andrea Vedaldi (in collab with @facebookai).
Paper: arxiv.org/abs/1910.08485
TorchRay: github.com/facebookresearch/…
20
69
(1/4) Meet VGGers @bmvc2019!
1. Self-supervised Video Representation Learning for Correspondence Flow, @LaiZihang, @WeidiXie
Poster: Session 1 #83 (Tuesday 14:00-16:15)
Oral: Video Analysis (Thursday 13:00-13:15)
Page: zlai0.github.io/CorrFlow/
1
19
70
Learning weakly symmetric 3D objects from single-view images only, without any additional supervision, eg. multi-views, keypoints, or prior shape models etc, by @elliottszwu, @chrirupp and Andrea Vedaldi.
Paper: arxiv.org/abs/1911.11130
Project: elliottwu.com/projects/unsup…
16
68
VGG Image Annotator (VIA) now support manual annotation of image, audio and video. VIA is a simple and standalone manual annotation software for image, audio and video. VIA runs in a web browser and does not require any installation or setup.
robots.ox.ac.uk/~vgg/softwar…
2
21
69
Congratulations to @subhc4 @irolaina @chrirupp and Andrea Vedaldi on receiving the Best Student Paper Award at @BMVCconf 🎉🎉
Check out the paper: "The Curious Layperson" robots.ox.ac.uk/~vgg/researc…
BEST STUDENT PAPER AWARD
The Curious Layperson: Fine-Grained Image Recognition without Expert Labels
Subhabrata Choudhury, Iro Laina, Christian Rupprecht and Andrea Vedaldi
University of Oxford
7
62
Check out 'RELATE': our new work on component-wise scene generation, in collaboration with @a2i_oxford, @uclcs, @adobe and @NianticLabs.
Disentangling foreground objects from background with inter-object relations modeling for generative model!
Website: geometry.cs.ucl.ac.uk/projec…
1
18
61
🤖 Do you have a PhD, and want to push the frontier of computer vision and robotics? 🤖
The Visual Geometry Group (VGG) in Oxford is hiring a postdoc! PI: Dr. João Henriques. Deadline: 2 June at noon (UK). More details:
eng.ox.ac.uk/jobs/job-detail…
14
60
10,163
We boost the performance of image retrieval systems by directly optimising a Smoothed Average Precision loss, Smooth-AP! By @AndrewB88429285 @WeidiXie @VickyKalogeiton and Andrew Zisserman.
website: robots.ox.ac.uk/~vgg/researc…
arXiv: arxiv.org/abs/2007.12163
#ECCV2020
2
17
58
🚨 Now Hiring 🚨
Are you a ML wizard, looking to work with cutting-edge research? 👀 VGG at Oxford wants YOU! 💼
👩🔬 Role: Postdoc Researcher
🌍 Location: University of Oxford
If you’ve got sharp code and sharper ideas, it’s to apply now!
🔗 Link👇
1
14
57
13,531
#CVPR2020 paper on uniting backpropagation saliency methods: There and Back Again: Revisiting Backpropagation Saliency Methods by
@SRebuffi, @ruthiecf, @xu__ji, Andrea Vedaldi.
Paper: arxiv.org/abs/2004.02866
Code: github.com/srebuffi/revisiti…
16
55
Excited to launch a web demo of our recently accepted #CVPR2020 paper!
Upload a single picture of human faces or cat faces, and visualize it in 3D using our unsupervised 3D object reconstruction model: robots.ox.ac.uk/~vgg/blog/un…
Learning weakly symmetric 3D objects from single-view images only, without any additional supervision, eg. multi-views, keypoints, or prior shape models etc, by @elliottszwu, @chrirupp and Andrea Vedaldi.
Paper: arxiv.org/abs/1911.11130
Project: elliottwu.com/projects/unsup…
17
52
Interested in CNN interpretability (or cute animal pics and visualizations)? Check out @ruthiecf and vedaldi's #CVPR18 paper (and code!) "Net2Vec: Quantifying and Explaining how Concepts are Encoded by Filters in Deep Neural Networks." openaccess.thecvf.com/conten…
1
22
44
Oriane Simeone from @Inria gave a fantastic talk today. Combing lightweight local features with activation maps to achieve dramatic gains for image retrieval!
1
13
47
#ICLR20 paper on novel category discovery @iclr_conf: Automatically Discovering and Learning New Visual Categories with Ranking Statistics by @Kai_Han_, @SRebuffi, Sebastien Ehrhardt, Andrea Vedaldi, Andrew Zisserman.
Code: github.com/k-han/AutoNovel
Page: robots.ox.ac.uk/~vgg/researc…
15
49
Watch the video recording of the presentation given by Andrea Vedaldi at #CVPR2020 Learning from Unlabeled Video Workshop:
Learning Representations and Geometry from Unlabelled Videos
piped.video/F2_SCM2Oqs4
15
49
Our recent @ICCV19 work on novel category discovery: Learning to Discover Novel Visual Categories via Deep Transfer Clustering by @Kai_Han_, Andrea Vedaldi, and Andrew Zisserman.
Paper: arxiv.org/abs/1908.09884
Code: github.com/k-han/DTC
Project page: robots.ox.ac.uk/~vgg/researc…
18
48
Congratulations to our former student David Novotny (@davnov134) and the team at @FacebookLondon on receiving the Honorable Mention Award @ICCV_2021!
Honourable Mention @ICCV_2021
Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstruction
Jeremy Reizenstein, Philipp Henzler, Roman Shapovalov, Luca Sbordone, Patrick Labatut, David Novotny (Facebook AI Research)
[Session 8 A/B]
2
47
We are happy to announce a new vacancy for a PDRA position in Computer Vision to work with Prof. Andrew Zisserman; more info here - eng.ox.ac.uk/jobs/job-detail… Application deadline is 17th March at 12 noon. Good luck!
12
50
7,519
We are excited to announce #VoxConverse, a large-scale, in the wild audio-visual speaker diarization dataset from YouTube, consisting of over 50 hours of video!
By Joon son Chung, @huh_jaesung, @NagraniArsha, @afourast & Andrew Zisserman.
3
19
48
Why do GANs generate such amazingly realistic images?
We are excited to have Antonio Torralba from @MIT_CSAIL @MITIBMLab here today, who presented recent findings in GAN dissection, which allows us to understand how GANs generate certain images and to manipulate the generation.
2
4
47
OxUVA - The largest dataset for training/testing long-term object tracking! (presented at #ECCV18, VGG+TVG+QUVA collaboration) #acronyms #objecttracking #dataset
oxuva.net
23
40
VGG twitter is now live! Follow @Oxford_VGG for updates on new papers, datasets, blog posts and other exciting research.
16
41
Congratulations Lukáš Neumann, Andrew Zisserman, and Andrea Vedaldi for winning the best paper award at the #CVPR19 workshop on "Precognition: Seeing through the Future" with their work on "Future Event Prediction: If and When".
Hyatt Shoreline A, 4.40 pm
robots.ox.ac.uk/~vgg/publica…
12
41
Max Bain presented his work “Count, Crop and Recognise: Fine-Grained Recognition in the Wild” at @ICCV19 CVWC workshop, joint work with @NagraniArsha, Andrew Zisserman.
Project page: robots.ox.ac.uk/~vgg/researc…
1
16
40
Check out our new work "Labelling unlabelled videos from scratch with multi-modal self-supervision" by @y_m_asano @mandelapatrick_ @chrirupp and Andrea Vedaldi in colab with @facebookai!
See below for our automatically discovered clusters on VGG-Sound! robots.ox.ac.uk/~vgg/researc…
1
10
39
Together with @MetaAI, we're organizing an #ECCV workshop on Neural Geometry and Rendering with an incredible line-up of distinguished speakers! @eccvconf
AND an exciting new Challenge on reconstructing common objects in 3D!
📢 ngr-co3d.github.io/
🏆 eval.ai/web/challenges/chall…
The Common Objects in 3D v2 #CO3Dv2 dataset is the next version of the first real-life dataset enabling large-scale 3D reconstruction of object categories. Together w/ the new version, we're also pleased to announce the CO3D challenge happening at #ECCV22! eval.ai/web/challenges/chall…
6
38
@erika_lu_ will be presenting #CVPR2021 Oral "Omnimatte: Associating Objects and Their Effects in Video"
⏰ Paper Session 4: Tue 8-10:30am PT / 11am-1:30pm ET / 4-6:30pm BST
🚩 omnimatte.github.io/
work w/ @forrestercole @talidekel A. Zisserman, W. Freeman, M. Rubinstein
Omnimatte: Associating Objects and Their Effects in Video
pdf: arxiv.org/pdf/2105.06993.pdf
abs: arxiv.org/abs/2105.06993
project page: omnimatte.github.io/
1
5
36
Introducing the NightOwls Pedestrian Detection Challenge @ICCV19! Detection at night is an under-represented yet very important problem with many interesting applications, while current state-of-the-art vision algorithms tend to perform badly.
nightowls-dataset.org/nighto…
18
36
Scalability in Autonomous Driving Workshop #CVPR2020 co-organised w/ @Waymo is happening on Monday afternoon June 15 (PDT)!
Join us with an exciting list of wonderful speakers @paul_m_newman @alexgkendall @kkitani @RaquelUrtasun @karpathy
sites.google.com/view/cvpr20…
#VGGatCVPR2020
10
37
📢 We are looking for a skilled and motivated Research Software Engineer to join the group. Applications are open until 14th September 2020.
Check more details and apply at: my.corehr.com/pls/uoxrecruit…
14
35
VoxCeleb: The largest audio-visual dataset of human interviews from YouTube, 7000+ identities, 1 million+ utterances & 2000+ hours of video: Urls, audio files, face detections, tracks and speaker meta-data available! #dataset #machinelearning #audiovisual robots.ox.ac.uk/~vgg/data/vo…
1
20
33
Congratulations to @gyunginshin, @WeidiXie and @SamuelAlbanie, who received the best paper award at the ILDAV workshop @ICCV2021 #ICCV2021!
Title: “All you need are a few pixels: semantic segmentation with PixelPick”
Paper: arxiv.org/pdf/2104.06394.pdf
Code: github.com/NoelShin/PixelPic…
2
8
31
📢 We are seeking 4 Senior Research Associates and 2 Postdoctoral Research Assistants in Computer Vision to join @Oxford_VGG or Biomedical Image Analysis Group, funded by @ERC_Research @EPSRC or Continental AG.
Details:
📌 my.corehr.com/pls/uoxrecruit…
📌 my.corehr.com/pls/uoxrecruit…
10
33
Excited to announce our #CVPR2020 paper in colab w/ @GoogleAI, "Speech2Action: Cross-modal Supervision for Action Recognition" is now online! By @NagraniArsha, Chen Sun, David Ross, @CordeliaSchmid, @rahulsukthankar and Andrew Zisserman. Check it out here: robots.ox.ac.uk/~vgg/researc…
13
33
Many congratulations to Prof. Andrea Vedaldi for winning one of the first Royal Society Faraday Discovery Fellowships from more than 600 initial applications. @royalsociety @oxengsci🥇🎉🙌🎉!!!!!! More info: ox.ac.uk/news/2025-08-28-exc…
4
33
3,475
Check out our new work on deep clustering:
LSD-C: Linearly Separable Deep Clusters, by @srebuffi, Sebastien Ehrhardt, @Kai_Han_, Andrea Vedaldi, Andrew Zisserman.
Code: github.com/srebuffi/lsd-clus…
Paper: arxiv.org/abs/2006.10039
6
29
Join our #CVPR2020 Workshop on The End-of-End-to-End A Video Understanding Pentathlon on Monday afternoon 15th June 2020 via Zoom, with great speakers @CordeliaSchmid @cvondrick and Anna Rohrbach.
robots.ox.ac.uk/~vgg/challen…
@SamuelAlbanie @NagraniArsha
#VGGatCVPR2020
We're excited to announce that the Video Pentathlon is now live!
Test out your video retrieval skills on five challenging benchmarks: MSRVTT, MSVD, YouCook2, ActivityNet and DiDeMo. More here: robots.ox.ac.uk/~vgg/challen…
Baselines and features provided! #CVPR2020 #video
10
31
Super excited to announce that former VGG member @jdthewlis is now the co-founder of @UnitaryAI! They aim to make the internet a better place using contextual video understanding. #EF #VideoUnderstanding
11
29
Excited to have Andrew Owens, Allan Jabri, Sheng-Yu Wang (@berkeley_ai), Alexander Berg (@unccs, @facebookai), @gulvarol (@Inria), Stefan Carlsson (KTH) and @AaronHertzmann (@AdobeResearch) to give amazing talks in our one-day visitors workshop!
#vggvisitors #computervision
5
29
Check out our #ECCV2020 spotlight work "MemDPC"!
We learn video representation by predicting the future with an external memory module in a self-supervised way.
by @TengdaHan, @WeidiXie and Andrew Zisserman
Arxiv: arxiv.org/abs/2008.01065
Code: github.com/TengdaHan/MemDPC
9
30
Congratulations to @e_kazakos, @NagraniArsha, Andrew Zisserman and @dimadamen for winning an Outstanding Paper Award at #ICASSP2021 @ieeeICASSP!
Details about the paper can be found here: ekazakos.github.io/auditorys…
"Slow-Fast Auditory Streams for Audio Recognition" by Evangelos Kazakos, Arsha Nagrani, Andrew Zisserman, & Dima Damen ieeexplore.ieee.org/document… #ICASSP2021
4
28
Annotate images quickly using the VGG Image Annotation (VIA) tool that runs in directly in a web browser (#opensource). Based solely on HTML, CSS and Javascript. Download at robots.ox.ac.uk/~vgg/softwar… #ai #annotationtool #machinelearning
1
15
27
Congratulations to @taeinkwon1 and the Team for presenting their work "A Dataset for Hand Pressure and Pose Estimation in Egocentric Vision" as a @CVPR
Highlight #CVPR25 🙌 🎉. yiming-zhao.github.io/EgoPre…
3
28
2,256
"Read and Attend: Temporal Localisation in Sign Language Videos" will be presented at #CVPR2021
⏰ Paper Session 12: Fri 3-5:30am PT / 6-8:30am ET / 11am-1:30pm BST / 12-2:30pm CET
🚩 robots.ox.ac.uk/~vgg/researc…
by @gulvarol, @LiliMomeni, @SamuelAlbanie, @afourast, Andrew Zisserman
5
25
Tomorrow @ICCV19 at 9:10am in Hall D1: @ruthiecf is presenting an oral on "Understanding Deep Networks via Extremal Perturbations and Smooth Masks" with
@mandelapatrick_ and Andrea Vedaldi (collab with @facebookai). #iccv2019 #XAI
arxiv.org/abs/1910.08485
github.com/facebookresearch/…
1
9
24
Watch the talk given by Abhishek Dutta on the VGG Image Annotator: piped.video/watch?v=WNgHy2hI…!
VIA: robots.ox.ac.uk/~vgg/softwar…
8
25
Two postdoc positions available in the VGG (apply by Feb 13th). Further infos here: eng.ox.ac.uk/jobs/job-detail…
16
25
New paper: Unsupervised Intuitive Physics from Past Experiences
by Sébastien Ehrhardt, Aron Monszpart, Niloy J. Mitra, Andrea Vedaldi
Learning to discover environment properties from past experiences compressed as dynamic images.
Code to be released soon.
arxiv.org/abs/1905.10793
6
21
'Night Owls: A Pedestrians at Night' #dataset presented by Lukáš Neumann #ACCV2018, in collaboration with @maxplanckpress and @drive_automated
It contains 279K frames & 40 sequences recorded across 3 countries!
Website: nightowls-dataset.org/
PDF: robots.ox.ac.uk/~vgg/publica…
1
14
24
We'd like to invite everyone to participate in the 2nd edition of NightOwls Detection Challenge #CVPR2020!
Object detection at night is an under-represented yet very important problem, with a lot of room for improvements.
More info: nightowls-dataset.org/nighto…
7
24
New paper: "Semi-Supervised Learning with Scarce Annotations"
by Sylvestre-Alvise Rebuffi, Sebastien Ehrhardt, Kai Han, Andrea Vedaldi, and Andrew Zisserman
Alternative optimisation method beating standard SOTA semi-supervised methods on transfer learning (10 labels per class)
1
7
18
Here is the paper: Local Features and Visual Words Emerge in Activations
arxiv.org/abs/1905.06358
#CVPR19
4
21
Not pictured, people sitting on the floor. @ludwigschubert from @openai gave an amazing talk on CNN interpretability and the progress of feature visualization techniques in the last few years. Thanks for visiting (and for coffee)!
2
2
21
"Capturing the Geometry of Object Categories from Video Supervision", by @davnov134 , @dlarlus and Andrea Vedaldi soon to appear in #TPAMI: Special Issue on The Best of #ICCV17 ! #computervision ieeexplore.ieee.org/document…
5
20
Excited to announce that the VoxCeleb journal paper is now online! It consists of all the information about the VoxCeleb1 and 2 datasets as well as new #speakerrecognition results! Check it out here: robots.ox.ac.uk/~vgg/publica…
11
21
Andrew's Bakerian Medal lecture on Computer Vision will be delivered on 3 May 2023 at the Royal Society.
Register for the event: ow.ly/IzQo50NvQa4
Livestreaming at: piped.video/c/royalsociety
Join us at the Royal Society this May to hear Professor Andrew Zisserman @Oxford_VGG, winner of the Bakerian Medal and Lecture 2023, deliver his prize lecture on the topic of #ComputerVision. Register now: ow.ly/IzQo50NvQa4
1
3
19
12,817
Recent work on self-supervised video representation learning by @TengdaHan, @WeidiXie and Andrew Zisserman
Our work on self-supervised video representation learning is online. State-of-the-art RGB-stream action classification accuracy, better than ImageNet pretrained weights!
Paper: arxiv.org/abs/1909.04656; Video: piped.video/43KIHUvHjB0; Code: github.com/TengdaHan/DPC @Oxford_VGG
5
21
Don't miss @aravindhm_ presenting his new work with @jdthewlis and Andrea Vedaldi at #ACCV2018 on Thursday 6th December (oral O5-7): "Cross Pixel Optical-Flow Similarity for Self-Supervised Learning"! robots.ox.ac.uk/~vgg/publica…
4
17
Tomas Yakub and Ankush Gupta presented their work at #NeurIPS2018 yesterday: '#UnsupervisedLearning of Object Landmarks through Conditional Image Generation' with Hakan Bilen and Andrea Vedaldi! Check out their webpage: robots.ox.ac.uk/~vgg/researc… for the paper, code & models!
9
21
We are excited to release a new large-scale audio-visual dataset VGG-Sound at #ICASSP2020!
by @Honglie_Chen, @WeidiXie, Andrea Vedaldi, Andrew Zisserman
Accessible via: robots.ox.ac.uk/~vgg/data/vg…
Presentation: 2020.ieeeicassp-virtual.org/…
#VGGSound
1
3
18
More audiovisual work! "Localizing Visual Sounds the Hard Way" will be presented at #CVPR2021
⏰ Friday, June 25 6:00 AM – 8:30 AM EDT / 11:00 AM – 1:30 PM BST
🚩 robots.ox.ac.uk/~vgg/researc…
by @Honglie_Chen @WeidiXie @afourast @NagraniArsha Andrea Vedaldi and Andrew Zisserman
Check out our CVPR21 paper 'Localizing Visual Sounds the Hard Way' with @WeidiXie, @afourast, @NagraniArsha, Andrea Vedaldi and Andrew Zisserman !
5
19
Congratulations to Professor Richard Hartley!
What if you actually could build Rome in a day? Prof Richard Hartley is one of the founders of Multiview Geometry, which establishes the construction of 3D models from sets of images or videos. He receives the Hannan Medal for his pioneering work.
20
2,659
The dataset we created for training unbiased networks is now released! 14,000 images across four ancestries to remove racial bias.
robots.ox.ac.uk/~vgg/data/la…
Find the paper at arxiv.org/abs/1809.02169
Awesome work by M. Alvi, @CNellaker, A. Zisserman
14
19
All four tracks for the VoxCeleb Speaker Recognition Challenge are now accepting submissions! There are open and closed training conditions for speaker verification, a track for self-supervised speaker verification and a new track for speaker diarisation! Deadline is 15th October
The submission server for our new diarisation track at the VoxCeleb Speaker Recognition Challenge (VoxSRC2020) is now open! This is the first time we are hosting this track so feedback is welcome. More details here: robots.ox.ac.uk/~vgg/data/vo… #VoxSRC2020
8
20
Luca is presenting their work today at 11 am, stand #5 @iclr2019
"Meta-learning with differentiable closed-form solvers"
For the problem of few-shot learning, they show how backpropagating through ridge regression is a simple way to achieve very fast adaptation at inference time.
1
3
18
We will anounce the winners at 2pm KST today at @ICCV19 7th Workshop on Computer Vision for Road Scene Understanding and Autonomous Driving!
sites.google.com/view/cvrsua…
Introducing the NightOwls Pedestrian Detection Challenge @ICCV19! Detection at night is an under-represented yet very important problem with many interesting applications, while current state-of-the-art vision algorithms tend to perform badly.
nightowls-dataset.org/nighto…
3
19
Yujie Zhong presenting his work "GhostVLAD for set-based #facerecognition" with @relja_work and Andrew Zisserman at #accv2018! The proposed network far exceeds the state of the art on the IJB-B face recognition dataset. Paper: arxiv.org/pdf/1810.09951.pdf
9
18
New work published in Science Advances! science.org/doi/10.1126/scia… Work by @maxhbain @NagraniArsha and AZ along with primatologists from multiple universities around the world!
Our work on Automated Audiovisual Behaviour Recognition in Wild Primates is finally out. An end-to-end detect, track and behaviour recognition pipeline, using both the audio and visual inputs (helpful for robustness in wild footage)
science.org/content/article/…
1
17
VGG papers at ECCV 2024 – Day 1 (Tuesday) 🚀
Come to talk to our members at the poster session🥳! 🧵👇
1
3
18
4,253
Congratulations Arsha on receiving the @ELLISforEurope PhD award!
Thank you and thanks to @ELLISforEurope! Congrats @thomaskipf :-)
2
18
Check out "Self-supervised Video Representation Learning by Pace Prediction" - new #ECCV2020 paper by Jiangliu Wang, Jianbo Jiao and Yun-Hui Liu.
Q&A session at 00:00 & 06:00 on Wed UK time.
Paper: arxiv.org/abs/2008.05861
Video: piped.video/wYHteK4BHlk
#VGGatECCV2020
10
19
Check your experiments in real-time with OverBoard - a native Python ML dashboard! github.com/jotaf98/overboard
12
18
New work at #ICASSP2020: ASR is All You Need: Cross-Modal Distillation for Lip Reading, by @afourast, Joon Son Chung, Andrew Zisserman.
Page: robots.ox.ac.uk/~vgg/researc…
Presentation: 2020.ieeeicassp-virtual.org/…
1
4
17
So good it made it to a t-shirt! Check out 'Semi-convolutional Operators for #InstanceSegmentation' by @davnov134, @SamAlbanie, @dlarlus and Andrea Vedaldi at #ECCV18 #MachineLearning robots.ox.ac.uk/~david/paper…
1
3
18
#ICLR2020 VGGers presenting exciting works @iclr_conf!
3. A critical analysis of self-supervision, or what we can learn from a single image @y_m_asano, @chrirupp, A. Vedaldi
Wed 12:00-14:00 & 17:00-19:00 GMT via iclr.cc/virtual/poster_B1esx…
openreview.net/forum?id=B1es…
1
6
16
'Gather-Excite'! Come to poster #61 at #NeurIPS2018, 5pm tomorrow (Wednesday) Room 210 & 230 AB #61 for some exciting new research by last years #ImageNet winners: Jie Hu, @ShirleyLishen, @SamAlbanie, Gang Sun and Andrea Vedaldi! arxiv.org/abs/1810.12348 #deepnetworks
7
16
The workshop is tomorrow (02/11) at 13:30, (not 14:00!) room 327A @ICCV19
Looking forward to hearing our invited speakers: David Forsyth, @zacharylipton, @WonderMicky, Oguzhan Gencoglu, and the authors of the contributed papers. Full program at preregister.vision.
#iccv2019
Tired of reviewers obsessing over state of the art results?
A new #ICCV2019 workshop is launching to trial the preregistration model in computer vision, i.e.:
Idea -> Paper -> Reviews -> Experiments
More info at preregister.vision!
5
16
Only 11 days to go to write up your idea (and how you'll evaluate it) and submit it to the ICCV'19 Preregistration Workshop!
preregister.vision
Tired of reviewers obsessing over state of the art results?
A new #ICCV2019 workshop is launching to trial the preregistration model in computer vision, i.e.:
Idea -> Paper -> Reviews -> Experiments
More info at preregister.vision!
8
17
Check out Bolt3D, a feed-forward 3D scene generation model ⚡️
⚡️ Introducing Bolt3D ⚡️
Bolt3D generates interactive 3D scenes in less than 7 seconds on a single GPU from one or more images.
It features a latent diffusion model that *directly* generates 3D Gaussians of seen and unseen regions, without any test time optimization.
🧵👇 (1/9)
1
11
2,691
#ICLR2020 VGGers presenting exciting works @iclr_conf!
1. Automatically Discovering and Learning New Visual Categories with Ranking Statistics @Kai_Han_ @SRebuffi S. Ehrhardt, A. Vedaldi, A. Zisserman
Mon 5-7 & 12-14 GMT via iclr.cc/virtual/poster_BJl2_…
robots.ox.ac.uk/~vgg/researc…
1
11
16