Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Inverse Image Frequency for Long-tailed Image Recognition

About

The long-tailed distribution is a common phenomenon in the real world. Extracted large scale image datasets inevitably demonstrate the long-tailed property and models trained with imbalanced data can obtain high performance for the over-represented categories, but struggle for the under-represented categories, leading to biased predictions and performance degradation. To address this challenge, we propose a novel de-biasing method named Inverse Image Frequency (IIF). IIF is a multiplicative margin adjustment transformation of the logits in the classification layer of a convolutional neural network. Our method achieves stronger performance than similar works and it is especially useful for downstream tasks such as long-tailed instance segmentation as it produces fewer false positive detections. Our extensive experiments show that IIF surpasses the state of the art on many long-tailed benchmarks such as ImageNet-LT, CIFAR-LT, Places-LT and LVIS, reaching 55.8% top-1 accuracy with ResNet50 on ImageNet-LT and 26.2% segmentation AP with MaskRCNN on LVIS. Code available at https://github.com/kostas1515/iif

Konstantinos Panagiotis Alexandridis, Shan Luo, Anh Nguyen, Jiankang Deng, Stefanos Zafeiriou• 2022

Related benchmarks

TaskDatasetResultRank
Object DetectionLVIS v1.0 (val)
APbbox25.8
518
Instance SegmentationLVIS v1.0 (val)
AP (Rare)18.6
189
Long-Tailed Image ClassificationPlaces-LT (test)
Accuracy40.2
61
Showing 3 of 3 rows

Other info

Follow for update