Mood Board + AI · Support Tools

Since completing my doctoral work on providing support tools for the creation of mood boards in 2009, the proliferation of online social media platforms such as Pinterest, Behance, or Dribbble dramatically changed the landscape of inspiration sources available for (digital) designers, affording online idea curation, portfolio creation, and self-promotion. However, very little is known about the potential effects that these online communities have on professional designers' inspiration-related practices and how designers view them.

We studied professional designers’ evolving inspiration-seeking practices. In particular, we ask how sources of inspiration might affect design thinking, especially with regard to inspirational material for creativity, reflection, and problem-solving. Once we updated our understanding of design practice, three computational support tools for the creation of mood boards were created (May AI?, Semantic Collage, Image Sense), each tackling a different key challenge.

Prototype: May AI?

Due to its quickly evolving and exploratory nature, providing computational support tools for early design process stages remains a challenge. The main hurdle for any machine-learning approach is to make creative contributions without insisting on systematic explicit feedback, thus distracting design thinking. It is also important to study methods that do not compromise the designer’s agency, allowing designers to work with rather than for an algorithm.

May AI? presents cooperative contextual bandits (CCB) that can learn to propose domain-relevant contributions and adapt their exploration/exploitation strategy. In the context of mood-board making, we developed a CCB for an interactive design ideation tool that suggests inspirational and situationally relevant materials, explores and exploits inspirational materials with the designer, and explains its suggestions to aid reflection.

Prototype: SemanticCollage

Designers often express and explore visual ideas through mood boards, which are collages composed of images, text, and object samples. Finding images that reflect abstract ideas is challenging and designers often spend significant time searching for the ‘right’ image. Unfortunately, current digital tools offer little support for searching beyond text queries and reflecting upon visual material.

SemanticCollage is a creativity support tool that supports designers' ideation and sense-making processes by exploiting state-of-the-art computer vision algorithms to enrich images with semantic labels.

Prototype: ImageSense

Given recent advances in machine learning (ML) and artificial intelligence (AI), we are interested in how to incorporate intelligent assistance into the ideation process, while leaving the human designer in control. The key challenge is how to support different levels of shared agency, between human designers and with the computer.

We present ImageSense, an intelligent, collaborative mood-board making tool that combines individual and shared work spaces, as well as collaboration with multiple forms of intelligent agents. In the collection phase, ImageSense offers fluid transitions between serendipitous discovery of curated images via ImageCascade, combined text- and image-based Semantic search, and intelligent AI suggestions for finding new images. For later composition and reflection, ImageSense provides semantic labels, generated color palettes, and multiple tag clouds to help communicate the intent of the mood board.

Workshops