Product Aesthetic Design: A Machine Learning Augmentation

TL;DR

This paper introduces a machine learning model that predicts aesthetic appeal and generates innovative automotive designs, significantly aiding the design process and improving consumer appeal with controllable, appealing outputs verified through empirical testing.

Contribution

The paper presents a novel hybrid generative model combining VAE, GAN, and supervised learning to predict and generate automotive aesthetics, validated with real-world data and extended to furniture design.

Findings

43.5% improvement over baseline in appeal prediction

Generated designs are appealing and resemble market trends

Model outperforms conventional ML and pretrained neural networks

Abstract

Aesthetics are critically important to market acceptance. In the automotive industry, an improved aesthetic design can boost sales by 30% or more. Firms invest heavily in designing and testing aesthetics. A single automotive "theme clinic" can cost over $100,000, and hundreds are conducted annually. We propose a model to augment the commonly-used aesthetic design process by predicting aesthetic scores and automatically generating innovative and appealing product designs. The model combines a probabilistic variational autoencoder (VAE) with adversarial components from generative adversarial networks (GAN) and a supervised learning component. We train and evaluate the model with data from an automotive partner-images of 203 SUVs evaluated by targeted consumers and 180,000 high-quality unrated images. Our model predicts well the appeal of new aesthetic designs-43.5% improvement relative to a uniform baseline and substantial improvement over conventional machine learning models and pretrained deep neural networks. New automotive designs are generated in a controllable manner for use by design teams. We empirically verify that automatically generated designs are (1) appealing to consumers and (2) resemble designs which were introduced to the market five years after our data were collected. We provide an additional proof-of-concept application using opensource images of dining room chairs.