# Interactive Image Manipulation with Natural Language Instruction Commands

## TL;DR

This paper introduces an interactive image manipulation system that uses natural language instructions to modify images, enabling more controllable and natural image editing through a neural network operating in latent space.

## Contribution

It presents a novel neural network framework that transforms source images into target images based on natural language instructions, enhancing controllability in image generation.

## Key findings

- Successfully generates target images from source images and instructions
- Demonstrates effective manipulation in a custom dataset
- Improves natural language conditioned image editing

## Abstract

We propose an interactive image-manipulation system with natural language instruction, which can generate a target image from a source image and an instruction that describes the difference between the source and the target image. The system makes it possible to modify a generated image interactively and make natural language conditioned image generation more controllable. We construct a neural network that handles image vectors in latent space to transform the source vector to the target vector by using the vector of instruction. The experimental results indicate that the proposed framework successfully generates the target image by using a source image and an instruction on manipulation in our dataset.