𝗗𝗮𝘆-𝟭𝟯𝟲 Computer Vision Learning 𝙇𝙞𝙗𝙧𝙖 𝙍-𝘾𝙉𝙉: Towards Balanced Learning for Object Detection by The Chinese University of Hong Kong and The University of Sydney Follow me for similar post : 🇮🇳 Ashish Patel Interesting Facts : 🔸 This is a paper in CVPR 2020 with over 6 citations. 🔸 It Outperforms with RetinaNet, Deformable R-FCN, SSD512, Mask-RCNN etc. ------------------------------------------------------------------- 𝗔𝗺𝗮𝘇𝗶𝗻𝗴 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 : https://lnkd.in/eCz48YQ Code : https://lnkd.in/e7CuYKX ------------------------------------------------------------------- 𝗜𝗠𝗣𝗢𝗥𝗧𝗔𝗡𝗖𝗘 🔸 Overview of the proposed Libra R-CNN: an overall balanced design for object detection which integrated three novel components (a) IoU-balanced sampling (b) balanced feature pyramid and (c) balanced L1 loss, respectively for reducing the imbalance at sample, feature, and objective level. #computervision #artificialintelligence #machinelearning
🇮🇳 Ashish Patel Bhaiya I wanna talk with you ..
For previous post visit this github : https://github.com/ashishpatel26/365-Days-Computer-Vision-Learning-Linkedin-Post
Cygnet.One•5K followers
4yThanks for sharing