Keep Calm and Carry On
开始阅读 Github
知
Python字符串神奇用法
1. 将一个字符串中的前两个’_’替换为’:’s = "a_b_c_d_e" s_replaced = s.replace("_", ":", 2) # 指定最大替换次数为2 print(s_replaced) 2. argparse命
阅读更多
DeepCache论文笔记
Title: DeepCache: Accelerating Diffusion Models for Free From NUS ArXiv 2023.12.07 Highlight 作者发现在diffusion生成过程中，相邻步的
阅读更多
CoDi论文笔记
Title: CoDi: Conditional Diffusion Distillation for Higher-Fidelity and Faster Image Generation From Google Research. C
阅读更多
Consistency Models论文笔记
Title: Consistency Models From OpenAI, Yang Song. ICML 2023. Highlight 所谓的Consistency Models一致性模型，是约束在同一条PF ODE轨迹上的任意
阅读更多
MAR 论文笔记
Title: Autoregressive Image Generation without Vector Quantization From MIT. ArXiv 2024.06.17 Highlight 作者认为自回归图像生成模型中
阅读更多
Clockwork Diffusion论文笔记
Title: Clockwork Diffusion: Efficient Generation With Model-Step Distillation From Qualcomm AI Research. CVPR 2024. H
阅读更多
DepthAnything论文笔记
Title: Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data From HKU & TikTok. CVPR 2024 Highlight 本
阅读更多
MUSE论文笔记
Title: Muse: Text-To-Image Generation via Masked Generative Transformers From Google Research. ICML 2023. Highlight 作
阅读更多
DiTs论文笔记
Title: Scalable Diffusion Models with Transformers From UC Berkeley and New York University ICCV 2023 (Oral Presentatio
阅读更多
SD3_Turbo论文笔记
Title: Fast High-Resolution Image Synthesis with Latent Adversarial Diffusion Distillation From Stability AI. ArXiv 202
阅读更多
Diff-Instruct论文笔记
Title: Diff-Instruct: A Universal Approach for Transferring Knowledge From Pre-trained Diffusion Models From PKU and No
阅读更多
Lumiere论文笔记
Title: Lumiere: A Space-Time Diffusion Model for Video Generation From Google ArXiv 2024.02.05 Highlight 背景：训练一个large
阅读更多
Sora技术报告解读
这里看的是OpenAI写的Sora技术报告 Sora是一个Video Generation生成器，可以生成不同时长、宽高比和分辨率的视频和图像，最多可达一分钟的高清视频。 Sora是使用不同长度的videos和images训练的tex
阅读更多
Preserving Fairness Generalization in Deepfake Detection 论文笔记
Title: 《Preserving Fairness Generalization in Deepfake Detection》 From Purdue University CVPR 2024 paper Highlight 文中提
阅读更多
DepthAnything V2论文笔记
Title: Depth Anything V2 From HKU & TikTok. ArXiv 2024.06.13 Highlight 本文从分析最近的两个深度估计的工作Depth Aynthing V1和Diffusio
阅读更多
SwiftBrush论文笔记
Title: SwiftBrush: One-Step Text-to-Image Diffusion Model with Variational Score Distillation From VinAI Research. CVPR
阅读更多
nanoGPT源码解析
这里看的是AK写的nanoGPT 1. wandb用法import wandb wandb.init(project=wandb_project, name=wandb_run_name, config=config) wandb.lo
阅读更多
TimeSformer论文笔记
Title: 《Is Space-Time Attention All You Need for Video Understanding?》 From Facebook ICML 2021 paper Highlight 背景：第一篇用
阅读更多
VAR论文笔记
Title: Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction From Bytedance Inc. ArXiv 20
阅读更多
VQGAN源码解析
这里看的是一个外国小哥复现的简单版本的VQGAN 1. tqdm的一个用法with tqdm(range(len(train_dataset))) as pbar: for i, imgs in zip(pbar, train_
阅读更多
SDXS论文笔记
Title: SDXS: Real-Time One-Step Latent Diffusion Models with Image Conditions From Xiaomi Inc. ArXiv 20240325 Highlig
阅读更多
动态规划专题
02/20/2024 - 01 300. 最长递增子序列 给定一个列表，求出最长递增子序列的长度。 class Solution: def lengthOfLIS(self, nums: List[int]) -> in
阅读更多
  我的梦想
人生只有一次，要过好我们自己的人生。—— David Chan
MUSE论文笔记 MUSE论文笔记
Title: Muse: Text-To-Image Generation via Masked Generative Transformers From Google Research. ICML 2023. Highlight 作
2024-07-06 Computer Vision
Image Generation Masked Generative Transformers
MAR 论文笔记 MAR 论文笔记
Title: Autoregressive Image Generation without Vector Quantization From MIT. ArXiv 2024.06.17 Highlight 作者认为自回归图像生成模型中
2024-06-22 Computer Vision
Diffusion Image Generation AutoRegressive
DepthAnything论文笔记 DepthAnything论文笔记
Title: Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data From HKU & TikTok. CVPR 2024 Highlight 本
2024-06-11 Computer Vision
Depth Estimation
DepthAnything V2论文笔记 DepthAnything V2论文笔记
Title: Depth Anything V2 From HKU & TikTok. ArXiv 2024.06.13 Highlight 本文从分析最近的两个深度估计的工作Depth Aynthing V1和Diffusio
2024-06-11 Computer Vision
Depth Estimation
Diff-Instruct论文笔记 Diff-Instruct论文笔记
Title: Diff-Instruct: A Universal Approach for Transferring Knowledge From Pre-trained Diffusion Models From PKU and No
2024-04-18 Image Generation
Diffusion Distillation
VAR论文笔记 VAR论文笔记
Title: Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction From Bytedance Inc. ArXiv 20
2024-04-07 Image Generation
Auto Regressive Model
SDXS论文笔记 SDXS论文笔记
Title: SDXS: Real-Time One-Step Latent Diffusion Models with Image Conditions From Xiaomi Inc. ArXiv 20240325 Highlig
2024-04-02 Image Generation
Diffusion Distillation Stable Diffusion
Clockwork Diffusion论文笔记 Clockwork Diffusion论文笔记
Title: Clockwork Diffusion: Efficient Generation With Model-Step Distillation From Qualcomm AI Research. CVPR 2024. H
2024-03-26 Image Generation
UNet Diffusion Distillation
SwiftBrush论文笔记 SwiftBrush论文笔记
Title: SwiftBrush: One-Step Text-to-Image Diffusion Model with Variational Score Distillation From VinAI Research. CVPR
2024-03-21 Image Generation
Diffusion Distillation Stable Diffusion
CoDi论文笔记 CoDi论文笔记
Title: CoDi: Conditional Diffusion Distillation for Higher-Fidelity and Faster Image Generation From Google Research. C
2024-03-17 Image Generation
Diffusion Distillation Consistency Models
SD3_Turbo论文笔记 SD3_Turbo论文笔记
Title: Fast High-Resolution Image Synthesis with Latent Adversarial Diffusion Distillation From Stability AI. ArXiv 202
2024-03-17 Image Generation
Diffusion Distillation Stable Diffusion
Consistency Models论文笔记 Consistency Models论文笔记
Title: Consistency Models From OpenAI, Yang Song. ICML 2023. Highlight 所谓的Consistency Models一致性模型，是约束在同一条PF ODE轨迹上的任意
2024-03-12 Image Generation
Diffusion Consistency Models Score-based model
1 / 2