The development of generative artificial intelligence (AI) enables large-scale product design automation. However, this automated process usually does not incorporate consumer preference information from the internal dataset of a company. Furthermore, external sources such as social media and user-generated content (UGC) websites often contain rich product design and consumer preference information, but such information is not utilized by companies when generating designs. We propose a semi-supervised deep generative framework that integrates consumer preferences and external data into the product design process, allowing companies to generate consumer-preferred designs in a cost-effective and scalable way. We train a predictor model to learn consumer preferences and use predicted popularity levels as additional input labels to guide the training procedure of a continuous conditional generative adversarial network (CcGAN). The CcGAN can be instructed to generate new designs with a certain popularity level, enabling companies to efficiently create consumer-preferred designs and save resources by avoiding the development and testing of unpopular designs. The framework also incorporates existing product designs and consumer preference information from external sources, which is particularly helpful for small or start-up companies that have limited internal data and face the "cold-start" problem. We apply the proposed framework to a real business setting by helping a large self-aided photography chain in China design new photo templates. We show that our proposed model performs well in terms of generating appealing template designs for the company.

翻译：生成式人工智能的发展使得大规模产品设计自动化成为可能。然而，这种自动化流程通常未能融入企业内部数据集中的消费者偏好信息。此外，社交媒体和用户生成内容网站等外部数据源常包含丰富的产品设计与消费者偏好信息，但企业在生成设计方案时并未充分利用这些信息。本文提出一种半监督深度生成框架，将消费者偏好与外部数据整合至产品设计流程中，使企业能够以经济高效且可扩展的方式生成符合消费者偏好的设计方案。我们训练了一个预测模型以学习消费者偏好，并将预测的受欢迎程度作为附加输入标签来指导连续条件生成对抗网络的训练过程。该网络可根据指定的受欢迎程度生成新的设计方案，使企业能够高效创建受消费者青睐的设计，并通过避免开发与测试不受欢迎的设计方案来节约资源。该框架还融合了来自外部数据源的现有产品设计与消费者偏好信息，这对于内部数据有限且面临"冷启动"问题的小型或初创企业尤为有益。我们将所提框架应用于实际商业场景，协助中国某大型自助摄影连锁企业设计新型照片模板。实验表明，所提模型在为企业生成具有吸引力的模板设计方面表现优异。