Image Enhanced Rotation Prediction for Self-Supervised Learning
The rotation prediction (Rotation) is a simple pretext-task for self-supervised learning (SSL), where models learn useful representations for target vision tasks by solving pretext-tasks. Although Rotation captures information of object shapes, it hardly captures information of textures. To tackle this problem, we introduce a novel pretext-task called image enhanced rotation prediction (IE-Rot) for SSL. IE-Rot simultaneously solves Rotation and another pretext-task based on image enhancement (e.g., sharpening and solarizing) while maintaining simplicity. Through the simultaneous prediction of rotation and image enhancement, models learn representations to capture the information of not only object shapes but also textures. Our experimental results show that IE-Rot models outperform Rotation on various standard benchmarks including ImageNet classification, PASCAL-VOC detection, and COCO detection/segmentation.
Code (0)
등록된 구현이 없습니다.
Tasks
Image EnhancementObjectPredictionSelf-Supervised LearningSimilar Papers 제목 키워드 기반
Self-Supervised Representation Learning by Rotation Feature Decoupling
We introduce a self-supervised learning method that focuses on beneficial properties of representation and their abilities in generalizing to real-world tasks. The method incorporates rotation invariance into the feature…
Representation LearningSelf-Supervised LearningDINO-RotateMatch: A Rotation-Aware Deep Framework for Robust Image Matching in Large-Scale 3D Reconstruction
This paper presents DINO-RotateMatch, a deep-learning framework designed to address the chal lenges of image matching in large-scale 3D reconstruction from unstructured Internet images. The method integrates a dataset-ad…
3D ReconstructionImage MatchingAggSS: An Aggregated Self-Supervised Approach for Class-Incremental Learning
This paper investigates the impact of self-supervised learning, specifically image rotations, on various class-incremental learning paradigms. Here, each image with a predefined rotation is considered as a new class for …
class-incremental learningClass Incremental LearningIncremental LearningSelf-Supervised LearningRotation-Equivariant Self-Supervised Method in Image Denoising
Self-supervised image denoising methods have garnered significant research attention in recent years, for this kind of method reduces the requirement of large training datasets. Compared to supervised methods, self-super…
DenoisingImage DenoisingExperimenting with Self-Supervision using Rotation Prediction for Image Captioning
Image captioning is a task in the field of Artificial Intelligence that merges between computer vision and natural language processing. It is responsible for generating legends that describe images, and has various appli…
DecoderImage CaptioningSelf-Supervised Learning