Generating Images Instead of Retrieving Them
| Autor: | Tuukka Ruotsalo, Antti Ukkonen, Pyry Joona |
|---|---|
| Přispěvatelé: | Department of Computer Science, Helsinki Institute for Information Technology |
| Rok vydání: | 2020 |
| Předmět: |
Matching (statistics)
Information retrieval Artificial neural network Computer science HEALTH INFORMATION ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Relevance feedback Information needs Medical information retrieval 02 engineering and technology 010501 environmental sciences 113 Computer and information sciences 01 natural sciences Image (mathematics) medical information retrieval symptom elicitation Adversarial system 020204 information systems 0202 electrical engineering electronic engineering information engineering Representation (mathematics) Symptom elicitation Generative grammar 0105 earth and related environmental sciences |
| Zdroj: | SIGIR |
| DOI: | 10.1145/3397271.3401129 |
| Popis: | Finding images matching a user's intention has been largely based on matching a representation of the user's information needs with an existing collection of images. For example, using an example image or a written query to express the information need and retrieving images that share similarities with the query or example image. However, such an approach is limited to retrieving only images that already exist in the underlying collection. Here, we present a methodology for generating images matching the user intention instead of retrieving them. The methodology utilizes a relevance feedback loop between a user and generative adversarial neural networks (GANs). GANs can generate novel photorealistic images which are initially not present in the underlying collection, but generated in response to user feedback. We report experiments (N=29) where participants generate images using four different domains and various search goals with textual and image targets. The results show that the generated images match the tasks and outperform images selected as baselines from a fixed image collection. Our results demonstrate that generating new information can be more useful for users than retrieving it from a collection of existing information. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|