当前位置:网站首页>Icml2022 | partial and asymmetric comparative learning of out of distribution detection in long tail recognition
Icml2022 | partial and asymmetric comparative learning of out of distribution detection in long tail recognition
2022-07-05 18:40:00 【Zhiyuan community】
Outside the existing distribution (OOD) The detection method is usually based on the training set with balanced class distribution . However , in application , Training sets usually have long tailed distributions . In this work , We first proved the existing OOD When the training set is a long tailed distribution, the performance of the detection method usually decreases significantly . Through analysis , We think this is because it is difficult for the model to compare a few tailed samples in the distribution with the real OOD Samples are distinguished , It makes the tail class easier to be incorrectly detected as OOD. To solve this problem , Partial and asymmetric supervised comparative learning (Partial and Asymmetric Supervised contrast Learning, PASCL), The model is explicitly encouraged to distinguish between tail intraclass distribution samples and OOD sample . In order to further improve the classification accuracy within the distribution , We propose auxiliary branch tuning (Auxiliary Branch Finetuning), It USES BN And two independent branches of the classification layer for anomaly detection and intra distribution classification . Intuition , Abnormal data and OOD Abnormal data has different underlying distributions . Our approach is CIFAR10-LT、CIFAR100-LT and ImageNet-LT False positive rate of abnormal detection on (FPR) Respectively 1.29%、1.45%、0.69%, The classification accuracy within the distribution is 3.24%、4.06%、7.89%. Code and pre trained models can be found in https://github.com/amazon-research/long-tailed-ood-detection find .
边栏推荐
猜你喜欢
随机推荐
vs2017 qt的各种坑
Vulnhub's darkhole_ two
2022年阿里Android高级面试题分享,2022阿里手淘Android面试题目
The 11th China cloud computing standards and Applications Conference | China cloud data has become the deputy leader unit of the cloud migration special group of the cloud computing standards working
Multithreading (I) processes and threads
写作写作写作写作
Writing writing writing
Copy the linked list with random pointer in the "Li Kou brush question plan"
Logical words in Articles
Use of websocket tool
兄弟组件进行传值(显示有先后顺序)
《ClickHouse原理解析与应用实践》读书笔记(5)
The easycvr platform reports an error "ID cannot be empty" through the interface editing channel. What is the reason?
Is it safe to make fund fixed investment on access letter?
Thoroughly understand why network i/o is blocked?
ConvMAE(2022-05)
Various pits of vs2017 QT
buuctf-pwn write-ups (9)
Pytorch yolov5 training custom data
【在优麒麟上使用Electron开发桌面应】