GaParmar/img2img-turbo

• highreadme#1

  Reposition the README's opening to emphasize one-step, turbo inference

  原因：

  当前

  # img2img-turbo
  
  **Paper** | **Sketch2Image Demo** 
  #### **Quick start:** [**Running Locally**](#getting-started) | [**Gradio (locally hosted)**](#gradio-demo) | [**Training**](#training-with-your-own-data)
  
  ### Cat Sketching
  <p align="left" >
  
  </p>
  
  ### Fish Sketching
  <p align="left">
  
  </p>
  
  We propose a general method for adapting a single-step diffusion model, such as SD-Turbo, to new tasks and domains through adversarial learning. This enables us to leverage the internal knowledge of pre-trained diffusion models while achieving efficient inference (e.g., for 512x512 images, 0.29 seconds on A6000 and 0.11 seconds on A100).

  复制粘贴的修复

  复制

  # img2img-turbo
  
  **One-step image-to-image translation with Stable Diffusion Turbo, achieving extreme speed and efficiency for tasks like sketch2image and day2night.**
  
  **Paper** | **Sketch2Image Demo** 
  #### **Quick start:** [**Running Locally**](#getting-started) | [**Gradio (locally hosted)**](#gradio-demo) | [**Training**](#training-with-your-own-data)
  
  ### Cat Sketching
  <p align="left" >
  
  </p>
  
  ### Fish Sketching
  <p align="left">
  
  </p>
  
  We propose a general method for adapting a single-step diffusion model, such as SD-Turbo, to new tasks and domains through adversarial learning. This enables us to leverage the internal knowledge of pre-trained diffusion models while achieving efficient inference (e.g., for 512x512 images, 0.29 seconds on A6000 and 0.11 seconds on A100).

  待办完成忽略
• mediumabout#2

  Add the arXiv paper link to the 'Homepage' field

  原因：

  复制粘贴的修复

  复制

  https://arxiv.org/abs/2403.12036

  待办完成忽略
• mediumtopics#3

  Add specific topics related to one-step inference and LCMs

  原因：

  当前

  computer-vision, deep-learning, generative-adversarial-network, generative-art, stable-diffusion

  复制粘贴的修复

  复制

  computer-vision, deep-learning, generative-adversarial-network, generative-art, stable-diffusion, one-step-inference, latent-consistency-models, image-to-image-translation, sd-turbo

  待办完成忽略

本次扫描解析到的品类 GEO 通道：google/gemini-2.5-flash, deepseek/deepseek-v4-flash

• 品类问题

  What are fast, one-step AI models for image-to-image translation, like sketch to photo?

  你：未被推荐
  AI 推荐顺序：

  1. Pix2Pix
  2. CycleGAN
  3. StyleGAN
  4. U-Net
  5. Stable Diffusion
  6. DALL-E 2
  7. Midjourney
  8. SPADE

  AI 推荐了 8 个替代方案，却始终没点名 GaParmar/img2img-turbo。这就是要补上的差距。

  查看 AI 完整回答
• 品类问题

  Seeking a deep learning solution for rapid image transformation, such as turning sketches into photos.

  你：未被推荐
  AI 推荐顺序：

  1. Pix2Pix
  2. CycleGAN
  3. SPADE
  4. StyleGAN
  5. ControlNet
  6. InstructPix2Pix

  AI 推荐了 6 个替代方案，却始终没点名 GaParmar/img2img-turbo。这就是要补上的差距。

  查看 AI 完整回答

• Metadata completeness

  warn

  建议：

• README presence

  pass

• Compared to common alternatives in this category, what is the core differentiator of GaParmar/img2img-turbo?

  pass
  ✓AI 明确点名了 GaParmar/img2img-turbo

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？

• If a team adopts GaParmar/img2img-turbo in production, what risks or prerequisites should they evaluate first?

  pass
  ✓AI 明确点名了 GaParmar/img2img-turbo

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？

• In one sentence, what problem does the repo GaParmar/img2img-turbo solve, and who is the primary audience?

  pass
  ✓AI 明确点名了 GaParmar/img2img-turbo

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？