전체 글319 [2025-2] 박지원 - GPTQ 논문) https://arxiv.org/abs/2210.17323 GPTQ: Accurate Post-Training Quantization for Generative Pre-trained TransformersGenerative Pre-trained Transformer models, known as GPT or OPT, set themselves apart through breakthrough performance across complex language modelling tasks, but also by their extremely high computational and storage costs. Specifically, due to their massarxiv.org 1. GPTQ란 GPTQ.. 2025. 7. 1. [2025-1] 최민서 - SDEDIT: Guided Image Synthesis and Editing with Stochastic Differential Equations [논문링크] https://arxiv.org/abs/2108.01073 SDEdit: Guided Image Synthesis and Editing with Stochastic Differential EquationsGuided image synthesis enables everyday users to create and edit photo-realistic images with minimum effort. The key challenge is balancing faithfulness to the user input (e.g., hand-drawn colored strokes) and realism of the synthesized image. Existing GANarxiv.org SDE Diffusi.. 2025. 6. 24. [2025-1] 이재호 - Diffusion Model Alignment Using Direct Preference Optimization https://arxiv.org/abs/2311.12908 - Bram Wallace et al, CVPR 2023 # Abstract 문제 인식:LLM은 RLHF로 사람의 선호에 맞게 정렬되지만, Diffusion Model은 아직 사람의 선호 학습이 널리 적용되지 않음.기존 접근:Text to image diffusion 모델에서는 고품질 이미지와 캡션으로 미세조정(fine-tuning)하는 방식이 일반적이었음.제안 방법:논문은 Diffusion-DPO라는 새로운 방법을 제안. 이는 **Direct Preference Optimization (DPO)**를 확산 모델에 맞게 변형하여, 사람이 선택한 이미지 쌍을 기반으로 직접 학습함. 1. Introduction 배경:Text-to-image diff.. 2025. 5. 31. [2025-1] 황징아이 - Convolutional Character Networks 논문 : https://arxiv.org/abs/1910.07954 Convolutional Character NetworksRecent progress has been made on developing a unified framework for joint text detection and recognition in natural images, but existing joint models were mostly built on two-stage framework by involving ROI pooling, which can degrade the performance on rearxiv.org 1. Introduction기존 Text Reading 모델은 2단계를 거친다텍스트 검출 (Text Detect.. 2025. 5. 31. 이전 1 2 3 4 5 6 7 8 ··· 80 다음
