전체 글(147)
-
[Generative] Python으로 FID 구하기
A. 수단- OS/Platform/Tool : Linux, Kubernetes(k8s), Docker, AWS- Package Manager : node.js, yarn, brew, - Compiler/Transpillar : React, Nvcc, gcc/g++, Babel, Flutter- Module Bundler : React, Webpack, ParcelB. 언어- C/C++, python, Javacsript, Typescript, Go-Lang, CUDA, Dart, HTML/CSSC. 라이브러리 및 프레임워크 및 SDK- OpenCV, OpenCL, FastAPI, PyTorch, Tensorflow, Nsight 1. What? (현상) FID(Frechet Inception Dista..
2024.10.15 -
[Generative] Class-Balancing Diffusion Models (CVPR’24)
Paper : https://openaccess.thecvf.com/content/CVPR2023/papers/Qin_Class-Balancing_Diffusion_Models_CVPR_2023_paper.pdfAuthorsAustin Univ. + Shanghai AI Lab, CVPR’24Main IdeaLong-tailed Dataset으로 DM을 학습할 때 생기는 문제점을 해결하기 위한 CBDM이라는 모델을 제안합니다.이는 Adjusted Transfer Probability로 구현되는데, 다시 말해 MSE형태의 추가적인 regularizer를 통해 구현됩니다.Tasks : T2I GenerationResults : CIFAR10, CIFAR100, CIFAR10LT, CIFAR100LT0. B..
2024.10.14 -
[Audio] Python Pedalboard를 활용해 Audio를 바꾸기
A. 수단- OS/Platform/Tool : Linux, Kubernetes(k8s), Docker, AWS- Package Manager : node.js, yarn, brew, - Compiler/Transpillar : React, Nvcc, gcc/g++, Babel, Flutter- Module Bundler : React, Webpack, ParcelB. 언어- C/C++, python, Javacsript, Typescript, Go-Lang, CUDA, Dart, HTML/CSSC. 라이브러리 및 프레임워크 및 SDK- OpenCV, OpenCL, FastAPI, PyTorch, Tensorflow, Nsight 1. What? (현상) 이번 글에서는 Pedalboard 패키지를 활용해..
2024.10.10 -
[Generative] A Recipe for watermarking Diffusion Models (ICLR’24)
Paper : https://arxiv.org/pdf/2303.10137AuthorsSea AI Lab +Tsinghua Univ, ICLR’24Main IdeaDiffusion Model을 위한 watermarking 방법을 정리하고, 분석한 논문입니다.Tasks : Unconditional Generation, T2I GenerationResults : FFHQ, AFHQv2, ImageNet-1K(64x64), CIFAR-10(32x32)0. Before Start... a. Defense Model b. Threat Model1. Problem2. Approach a. Unconditional & Class-conditional Generation b. Text-To-Image ..
2024.10.07 -
[Generative] Shap·E: Generating Conditional 3D Implicit Functions (arxiv'23)
Paper : https://arxiv.org/pdf/2305.02463AuthorsOpenAI, arxiv’23Main Idea기존과 다르게 NeRF와 Textured Mesh 두가지 모두로 render되는 implicit function 파라미터를 생성합니다.기존 Point-E 모델에 비해 빠르게 Converge하면서도 좋은 퀄리티의 모델링이 가능합니다.Tasks : Text-To-3D Object GenerationResults : Rendered From RGBA data0. Before Start... a. INR(Implicit Neural Representations) b. 3D Representation with SDF c. Point-E1. Problem 2. Approac..
2024.09.21 -
[Markdown] README 템플릿
A. 수단- OS/Platform/Tool : Linux, Kubernetes(k8s), Docker, AWS- Package Manager : node.js, yarn, brew, - Compiler/Transpillar : React, Nvcc, gcc/g++, Babel, Flutter- Module Bundler : React, Webpack, ParcelB. 언어- C/C++, python, Javacsript, Typescript, Go-Lang, CUDA, Dart, HTML/CSSC. 라이브러리 및 프레임워크 및 SDK- OpenCV, OpenCL, FastAPI, PyTorch, Tensorflow, Nsight 1. What? (현상) README.md는 Github의 프로젝트를 설명..
2024.09.19