Annotate Research Papers: ML/DL 논문을 '쉽게' 읽을 수 있는, 주석이 포함된 논문 📖 저장소

9bow · 9월 28, 2023, 12:31오전

ML/DL 논문은 유용하지만 읽기 어렵습니다
~~(한때 불면증으로 고생했는데, 같은 팀 박사님 께서는 논문 읽기를 권하시더라고요 )~~

새로 나오는 논문은 기하급수로 증가하고, 제가 읽는 속도는 느린 산술급수급이라,
Paper Queue;;가 계속 늘어만 가고 있었는데;;

조금이나마 도움이 될만한 저장소를 TLDR 뉴스레터에서 발견하여 가져옵니다.
(아직 양이 많지는 않아서 주로 처음 보는 분야의 논문을 읽을 때 유용하지 않을까... 싶습니다. )

papers

지금까지의 논문 목록들은 아래와 같습니다.

Field	Category	Annotated Paper
Computer Vision		Adaptive Risk Minimization		Abstract
		Axial DeepLab	Code	Abstract
		ConvNext	Code	Abstract
		EfficientNetsV2	Code	Abstract
	Supervised	Flow-edge Guided Video Completion	Code	Abstract
		Is Batch Norm Unique?		Abstract
		Knowledge Distillation: A good teacher is patient and consistent	Code	Abstract
		RandConv	Code	Abstract
		Polyloss	Code	Abstract
		Scaling Down Deep Learning	Code	Abstract
		Segment Anything		Abstract
		Supervised Contrastive Learning	Code	Abstract
		Vision Transformer	Code	Abstract




		Are all negatives created equal in contrastive instance discrimination?		Abstract
		Towards Domain-Agnostic Contrastive Learning		Abstract
	Self-Supervised	Emerging Properties in Self-Supervised Vision Transformers	Code	Abstract
		Decoder Denoising Pretraining		Abstract
		Masked Autoencoders	Code	Abstract
		Swav	Code	Abstract
		What Should Not Be Contrastive in Contrastive Learning		Abstract


	Semi-Supervised	CoMatch	Code	Abstract




Diffusion Models		Understanding Diffusion Models		Abstract
		On the Importance of Noise Scheduling for Diffusion Models		Abstract
		Emergent Correspondence from Diffusion Models		Abstract


GANs		CycleGan	Code	Abstract




Interpretability and Explainability		What is being transferred in transfer learning?	Code	Abstract
		Explaining in Style	Code	Abstract




NLP		Do Language Embeddings Capture Scales?		Abstract
		mSLAM		Abstract
		Cramming		Abstract




Speech		SpeechStew		Abstract
		mSLAM		Abstract
		WhisperX	Code	Abstract




Others		Multi-Task Self-Training for Learning General Representations		Abstract
		Decoder Denoising Pretraining for Semantic Segmentation		Abstract
		Sigmoid Loss for Image-Text Pretraining		Abstract

아래 저장소에 계속 업데이트 한다고 하니, 관심 있으신 분들께서는 Watch 나 Star 해두시는 것을 추천합니다.