Edinburgh Vision

Check out this upcoming ICCV workshop out-of-distribution generalization, co-organized by @BingchenZhao and team.

📢📢📢 Job Opportunity! Join Dr Elliot J. Crowley and the Bayesian and Neural Systems Group as a postdoc in NAS and AutoML at the University of Edinburgh. 24-month funded position starting Autumn 2023. PhD in ML or related field required. Apply now: elxw.fa.em3.oraclecloud.com/hcmUI/Candidat…

Meta Omnium is a multi-task few-shot learning benchmark to evaluate generalization across CV tasks. Work by @OBohdal @EricTian1102, @yongshuozong, @chavhan_ruchika, @dali_academic, @henrygouk, @tmh31, will be presented on 20/6 afternoon. Project page: edi-meta-learning.github.io/meta-omnium/

Our Gen Li will present his paper at @CVPR on the 21st (morning session). This paper learns how humans interact with various objects in a weakly-supervised manner, transferring knowledge to static objects to ground actionable regions. Project page: reagan1311.github.io/locate/

Are you interested in doing a PhD that combines machine learning with remote sensing? Consider applying for eo-cdt.org/projects/measu…. This is a fully-funded PhD studentship at Leeds where you will be working with Professor Andy Challinor and Dr Elliot Crowley from Edinburgh

We are organising a workshop in the BMVC. We have a nice line up with interesting talks starting at 10.00 today. See sites.google.com/view/universal…. If you are attending #BMVC2022 , our workshop is in Jardine Suite (Ground Floor).

Octave Mariotti will be presenting ViewNeRF, an unsupervised method to recover category level cameras using radiance fields at BMVC this afternoon. Drop by poster 740 for a chat with him!

We present another poster today at #BMVC2022 today (2-3:30pm), called "An Action Is Worth Multiple Words: Handling Ambiguity in Action Recognition", presenting new challenge in action labelling, especially in #EpicKitchens egocentric dataset. @davmoltisanti @oisinmacaodha

We are presenting our work today (11/23) at #BMVC2022 from 2:00 PM - 3:30 PM! If you are at BMVC and interested in temporal segmentation of steps in instructional videos, stop by our poster for a chat!

Today we're presenting "Capturing Temporal Information in a Single Frame: Channel Sampling Strageties for Action Recognition" at #BMVC2022 in London. Please drop by our poster between 2-3:30 PM! github.com/kiyoon/channel…

Today (22/11) we present "SVL-Adapter: Self-Supervised Adapter for Vision-Language Pretrained Models" at #BMVC2022 in London. If you are at BMVC and interested to learn how SSL can help large VL models adapt to challenging visual tasks drop by our poster between 2-3:30pm!

Today (10/26) we're presenting our work on Fine-Grained Audiovisual Categorization with the SSW60 Dataset at #ECCV2022 poster #115 from 3:30-5:30pm. What are the benefits of image, audio, and video modalities, and how should you train your model? Drop by and find out! 🐦🤖🐤

BRACE, our new dataset for dance motion synthesis will be presented this morning at #ECCV2022. Come and have a look at stand 59!

Check out @eli_cole_ poster this afternoon at #ECCV2022

Our own Davide Moltisanti @davmoltisanti is one of the outstanding reviewers of ECCV 2022! @eccvconf eccv2022.ecva.net/program/outsta…

Can we use self-supervised learning to better adapt vision-language models to real-world challenging tasks? Check our #BMVC2022 paper "SVL-Adapter: Self-Supervised Adapter for Vision-Language Pretrained Models" to learn more. paper: arxiv.org/abs/2210.03794 1/8

An action is worth multiple words. Check out our BMVC 22 paper, where we study semantic ambiguity in action recognition. arxiv.org/abs/2210.04933. With Kiyoon Yim (first author), Oisin Mac Aodha @oisinmacaodha and Laura Sevilla

Excited to share my new work "A Closer Look at Temporal Ordering in the Segmentation of Instructional Videos" accepted at #BMVC2022 👇 arxiv.org/abs/2209.15501 Thanks to @frank_e_keller, Laura Sevilla and Shreyank Gowda. Looking forward to present it in London #BMVC2022

Check out Kiyoon's @KiyoonKim7 BMVC 22 paper, where they show various video channel sampling strategies to capture longer temporal information. arxiv.org/abs/2201.10394. With Shreyank Gowda, Oisin Mac Aodha @oisinmacaodha and Laura Sevilla. Code at github.com/kiyoon/channel…

We are organizing the Universal Representations for Computer Vision Workshop at #BMVC2022 on Nov. 24th in London. We have a fantastic set of speakers! #h.ajo7i1l7kcqi" target="_blank" rel="nofollow noopener">sites.google.com/view/universal…… We invite submissions of regular and short papers (See Call for Paper for more details).