Our new work (DGC-Net) regarding dense pixel correspondences and image alignment is available online arxiv.org/abs/1810.08393 w/ Viktor Tiulpin 🇺🇦 Marc Pollefeys #ComputerVision #Robotics #DeepLearning

I will give a talk next week at "The 2nd International Workshop on Big Data Analytic for Cyber Crime Investigation" in Seattle on the application of #DeepLearning to #forensics based on a solution to a Kaggle challenge. The written version of it medium.com/@iglovikov/for…

2018 was the year of @MLflowOrg. I wonder if 2019 will be the year of dvc.org. Interesting convergence. See introduction by Dmitry Petrov in youtube.com/watch?v=BneW7j… #data #DataScience #DataEngineering #pipelines #software #BigData #MachineLearning #mlflow #dvc

It's a rewarding job to work on DVC.org and hear Rachael Tatman @[email protected] talk about it! Great summary ➡ kaggle.com/rtatman/kernel… We are excited to guide more Kaggle folks🧙‍♀️🧙‍♂️ through using DVC. Let us know your pains and thoughts to make it even better🚀💥

После 18-го февраля начнется новая версия курса "Deep Learning на пальцах", на этот раз открытая для всех. Лекции, где можно задавать вопросы, задания на Питоне, тусовка в чатиках. Если вы все откладывали разобраться с deep learning - самое время! dlcourse.ai

Applications are finally open! The school will take place in Moscow, August 20–25. Applications close on April 15, 23:59 Moscow Time (UTC+3) deepbayes.ru/#apply

kaggle.com/artgor/cancer-… Ever wished for an easier way to train Pytorch models? To use lr finder or one cycle without fastai easily? Kekas (github.com/belskikh/kekas) is a higher level library which helps with these things. I wrote a kerne with this library as an example!

I would like to announce that @albuslaev creaf Одна інтеллектуальна одиниця Vladimir Iglovikov are releasing image augmentation library #albumentations v0.2.0. The new features are key point transformations, multiple targets of the same type and many other improvements. github.com/albu/albumenta…

Do you know what's in common between ITALY, BRAZIL, QUEBEC, CRIMEA, CUBA, IRAN, SYRIA, NORTH KOREA, SUDAN, RUSSIA, UKRAINE, KAZAKHSTAN, and BELARUS? Residents of all these countries are prohibited from participating in #PoweredByTF 2.0 Challenge. But why???

Importance Weighted Hierarchical Variational Inference is online. More expressive neural samplers for Variational Inference with a novel multisample variational upper bound on log q(z|x). Paper: arxiv.org/abs/1905.03290 Talk: youtu.be/pdSu7XfGhHw artem.sobolev.name/posts/2019-05-…

Glad to announce my new paper on #breastcancer Cellularity Assessment using #deeplearning! ArXiv preprint lnkd.in/ddQf6S2 Joint work with Viktor Alexandr Kalinin Alexey Shvets Vladimir Iglovikov Sergey Nikolenko #breastcancer #deep_learning #AI #machine_learning

We willl be presenting our work "Divide and Conquer the Embedding Space for Metric Learning" at #CVPR2019 on Tuesday 18th: Poster 24 at 10:15. Authors: Me, Vadim Tschernezki, Uta Büchler (Uta Büchler) and Björn Ommer. Paper and Code: bit.ly/dcesml

You end up with the following awful practices. #1 Copying & pasting your article in your poster.

On July 16, 1945 the Trinity test became the first nuclear detonation in human history. It would be followed over the next 74 years by more than 2000 more.

Preparing for a #MachineLearning #DataScience interview? One retweet - one theoretical interview question in the thread 👇 Feel free to give your answers Let's start! #100DaysOfCode #100DaysOfMLCode

You must have seen a lot of mask detection videos recently. Now learn from the expert, Vladimir Iglovikov in the next session of Talks! Google Calendar: bit.ly/3igXCBy Youtube: youtube.com/watch?v=VAg8pM… 14th Aug. 1800h CET / 2130h IST / 0900h PDT LIKE.SUBSCRIBE.SHARE.

Great advice by Vladimir Iglovikov on sharing work. youtube.com/watch?v=VAg8pM…

Blog post on how we rented a place at Airbnb for thanksgiving. TLDR: We got burgled. Details in the text. ternaus.blog/incident/2020/…

Kandinsky 2.1 inherits best practicies from Dall-E 2 and Latent diffucion, while introducing some new ideas. As text and image encoder it uses CLIP model and diffusion image prior (mapping) between latent spaces of CLIP modalities. This approach increases the visual performance

Come and talk with the authors about a new instruction benchmark, MERA, at #ACL2024NLP. 📍Poster Session 5 📅August 13, ⏰16:00–17:30 👉Paper: aclanthology.org/2024.acl-long.… 👉Repo: github.com/ai-forever/MERA