MUHAMMAD SHER AFGAN

代表作品 / 项目

• Fine-Tuning-the-Image-Encoder-of-clip-using-pre-Trained-CLIP-ViT-Large-Patch14⭐ 2Optimize CLIP-ViT-Large-Patch14.ipynb with our tailored image encoder fine-tuning script. Quickly adapt the model to your needs for enhanced performance on image-based tasks.
• Generative-modeling-for-beginners-AE-VAE-VQAE-and-GAN-⭐ 2Implementation with easy understanding of linear Autoencoder, Variational Autoencoder, Convolutional Variational Autoencoder and Generative adversarial network using PyTorch to generate Artificial new images. All the work done with the generation perspective has been related to image-based data.
• CLIP-based-People-Photo-Classifier⭐ 0Harness the power of CLIP (Contrastive Language-Image Pre-training) to classify photos of people with our Python code implementation. This repository provides a seamless integration of CLIP for accurate and context-aware classification of images featuring individuals.
• Conditional-Diffusion-for-MNIST-Code-Implementation⭐ 0Explore conditional image generation with our Python code implementation, "Conditional_Diffusion_MNIST-main." This repository leverages diffusion models to generate diverse and conditioned images within the MNIST framework. Perfect for researchers and enthusiasts in generative modeling and image synthesis.
• CycleGAN-for-Medical-Image-Style-Transfer-to-Natural-Images⭐ 0Seamlessly transfer medical image styles to natural aesthetics with our concise CycleGAN implementation. Effortlessly enhance visualizations and adapt medical imagery for diverse contexts.