@IrafmVision profile picture
IRAFM-Vision @IrafmVision

We are a group of academic image processing experts developing own theory and methods with industry applications

Joined September 2018
  • Tweets 37
  • Following 10
  • Followers 19
14 Photos and videos
Náš kolega založil YouTube kanál, kde vysvětluje laikům umělou inteligenci. Jestli se chcete něco zajímavého dozvědět, dejte si odběr, ať vám neuniknou nová videa youtube.com/playlist?list=PL…

Umělá inteligence s Petrem

youtube.com
1
We have made a free, simple tool that easily produces selective colored images; see the demonstration clip: youtube.com/watch?v=dlygaNzZ…
1
Thank you, @signatelab, for such a nice Christmas gift.
3
Hi, @nmygle how about joining knowledge?
株式会社SIGNATE【公式】@signatelab
9 Dec 2020
【新着コンペ🏞️防衛装備庁　オフロード画像のセグメンテーションチャレンジ】 2020/12/9 - 2021/2/15 オフロード環境等の実走行で得られた画像を領域分割するアルゴリズムを作成し、認識精度と推論速度を競うコンテストを開催します！ みなさまの挑戦お待ちしています！！ signate.jp/competitions/101
1
2
Making of an online interview for television about the 4th Tellus competition. @signatelab
4
Ostravska univerzita @Ostravska_Uni zanechala stopu ve vesmiru! V celosvetove soutezi na vyvoj algoritmu pro automatizovane zpracovani snimku porizovanych z druzice jsme skoncili druzi v konkurenci 803 prihlasenych lidi a ziskavame tak krasnou financni vyhru. @signatelab
2
16
#Signate competition on processing data from satellite. It seems that team-up with a Japanese researcher was a perfect idea. Now we need to keep the position for 29 days until the end of the competition.
6
We are finishing a paper to #ICLR dealing with a real-time refinement of BB predictions. The refinement can increase the mAP of SOTA algorithms (confirmed for #FasterRCNN, #RetinaNet, #DETR), but it can also refine GT labels to be more precise. See the image of refined #COCO GT.
1
1
A new #Signate competition is here! The goal is to segment coastlines semantically from #Tellus satellite images. We have a nice start, but 75 days left.
1
2
Wow, a big step ahead for AI.
Raphaël Millière@raphaelmilliere
31 Jul 2020
I asked GPT-3 to write a response to the philosophical essays written about it by @DrZimmermann, @rinireg @ShannonVallor, @add_hawk, @AmandaAskell, @dioscuri, David Chalmers, Carlos Montemayor, and Justin Khoo published yesterday by @DailyNousEditor. It's quite remarkable!
1
We have published realtime inferences of Poly-YOLO lite on webcam. The model is pre-trained on COCO dataset and runs 40fps without further optimization. youtube.com/watch?v=MP4MfHpU…

Poly-YOLO lite webcam pretrained on COCO dataset

Model resolution 416x416px, realtime inference.Repository: https://gitlab.com/irafm-ai/poly-yoloPaper: https://arxiv.org/abs/2005.13243Tutorial: https://medi...

youtube.com
1
We just published Tutorial on Poly-YOLO: Bounding boxes link.medium.com/BlO9esT1b8

Tutorial on Poly-YOLO: Bounding boxes

In this article, I would like to explain details about our new CNN network called Poly-YOLO, its principles and how to use it in a step by…

link.medium.com
IRAFM-Vision retweeted
Suzana Ilić
@suzatweet
9 Jul 2020
Huge congrats to the @__MLT__ team @yoovraj, @alisher_ai, @knaveen, @hjmkt3126 and Ben for winning the gold medal and 3rd place in the Edge AI competition organized by Japanese Ministry of Economy, Trade and Industry and NEDO! 👏👏👏 @signatelab signate.jp/competitions/256/…
2
7
36
Our team is a runner-up in #Signate competition with their #POLYYOLO approach. signate.jp/competitions/256 gitlab.com/irafm-ai/signate_…
2
We just published Poly-YOLO: higher speed, more precise detection and instance segmentation for YOLOv3 link.medium.com/S9ayUecZQ6

Poly-YOLO: higher speed, more precise detection and instance segmentation for YOLOv3

Every idea needs a Medium

link.medium.com
Do you know about any better way how to spend Valentine's morning than listening to a super cool talk? Correct, you do not! Do you wanna hear about Freddie Mercury and Michael Faraday within one presentation? Correct, you do! You are invited! facebook.com/events/11472122…

Pavel Vlášánek - Seminar on fuzzy modeling

Other event in Ostrava, Czech Republic by Institute for Research and Applications of Fuzzy Modeling - IRAFM on Thursday, February 14 2019

facebook.com
We are presenting a new image sharpness measure at SCIS&ISIS2019 #conference, #Japan.
Our hobby is text detection and recognition in difficult conditions. In realtime. #OCR #machineLearning
1
3
Our latest paper co-authored by Pavel Vlasanek and Irina Perfilieva about #image #inpainting has been published in #Springer just today 😎 rdcu.be/9WU0
Load more