aniketdiffusion

Diffusion Classifier leverages pretrained diffusion models to perform zero-shot classification without additional training

[CVPR 2024] Code release for "InstanceDiffusion: Instance-level Control for Image Generation"

This repository is the official implementation of Disentangling Writer and Character Styles for Handwriting Generation (CVPR23).

Text-To-Image Generation with Chinese Characters

[ICCV2023] Self-supervised Character-to-Character Distillation for Text Recognition

Experiment with diffusion models that you can run on your local jupyter instances

Official Pytorch Implementation of DenseDiffusion (ICCV 2023)

Main repository for "CharacterBERT: Reconciling ELMo and BERT for Word-Level Open-Vocabulary Representations From Characters"

This is the official code for ``Can Diffusion Model Achieve Better Performance in Text Generation? Bridging the Gap between Training and Inference!`` (ACL 2023 Findings)

A novel method that provides greater control over generated images by guiding the internal representations of the pre-trained Stable Diffusion.

ACM Multimedia 2023: DocDiff: Document Enhancement via Residual Diffusion Models. Also contains 1597 red seals in Chinese scenes, along with their corresponding binary masks.

Bring portraits to life!

Code for project DS-Fusion

Official Implementation for "Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image Diffusion Models" (SIGGRAPH 2023)