# Interactive Optimization of Generative Image Modeling using Sequential Subspace Search and Content-based Guidance

## TL;DR

This paper introduces a human-in-the-loop optimization system for generative image models that enables users to explore and refine images interactively without needing specialized architectures or additional data.

## Contribution

It presents a versatile, model-agnostic interactive framework for exploring latent spaces and guiding image generation through user input and editing tools.

## Key findings

- System enables direct latent space exploration and editing.
- Demonstrates superior user satisfaction in comparative studies.
- Applicable to various pre-trained generative models.

## Abstract

Generative image modeling techniques such as GAN demonstrate highly convincing image generation result. However, user interaction is often necessary to obtain the desired results. Existing attempts add interactivity but require either tailored architectures or extra data. We present a human-in-the-optimization method that allows users to directly explore and search the latent vector space of generative image modeling. Our system provides multiple candidates by sampling the latent vector space, and the user selects the best blending weights within the subspace using multiple sliders. In addition, the user can express their intention through image editing tools. The system samples latent vectors based on inputs and presents new candidates to the user iteratively. An advantage of our formulation is that one can apply our method to arbitrary pre-trained model without developing specialized architecture or data. We demonstrate our method with various generative image modeling applications, and show superior performance in a comparative user study with prior art iGAN.