CVDec 11, 2023

InstructAny2Pix: Flexible Visual Editing via Multimodal Instruction Following

arXiv:2312.06738v423 citationsh-index: 7Has Code
Originality Incremental advance
AI Analysis

This addresses the need for more flexible and natural multi-modal control in computer vision applications, though it appears incremental by combining existing techniques like multi-modal encoders, diffusion models, and LLMs.

The paper tackles the problem of providing fine-grained control for visual image editing by introducing InstructAny2Pix, a system that allows users to edit images using instructions involving audio, images, and text, achieving novel instruction-guided editing tasks.

The ability to provide fine-grained control for generating and editing visual imagery has profound implications for computer vision and its applications. Previous works have explored extending controllability in two directions: instruction tuning with text-based prompts and multi-modal conditioning. However, these works make one or more unnatural assumptions on the number and/or type of modality inputs used to express controllability. We propose InstructAny2Pix, a flexible multi-modal instruction-following system that enables users to edit an input image using instructions involving audio, images, and text. InstructAny2Pix consists of three building blocks that facilitate this capability: a multi-modal encoder that encodes different modalities such as images and audio into a unified latent space, a diffusion model that learns to decode representations in this latent space into images, and a multi-modal LLM that can understand instructions involving multiple images and audio pieces and generate a conditional embedding of the desired output, which can be used by the diffusion decoder. Additionally, to facilitate training efficiency and improve generation quality, we include an additional refinement prior module that enhances the visual quality of LLM outputs. These designs are critical to the performance of our system. We demonstrate that our system can perform a series of novel instruction-guided editing tasks. The code is available at https://github.com/jacklishufan/InstructAny2Pix.git

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes