File Download
Supplementary
-
Citations:
- Appears in Collections:
Conference Paper: Lightweight Generative Adversarial Networks for Text-Guided Image Manipulation
Title | Lightweight Generative Adversarial Networks for Text-Guided Image Manipulation |
---|---|
Authors | |
Issue Date | 2020 |
Publisher | Neural Information Processing Systems Foundation, Inc. The Journal's web site is located at https://papers.nips.cc/ |
Citation | 34th Conference on Neural Information Processing (NeurIPS), Virtual Conference, 6-12 December 2020. In Larochelle, H ... et al (eds.), Advances in Neural Information Processing Systems 33 (NeurIPS 2020) Proceedings How to Cite? |
Abstract | We propose a novel lightweight generative adversarial network for efficient image manipulation using natural language descriptions. To achieve this, a new word-level discriminator is proposed, which provides the generator with fine-grained training feedback at word-level, to facilitate training a lightweight generator that has a small number of parameters, but can still correctly focus on specific visual attributes of an image, and then edit them without affecting other contents that are not described in the text. Furthermore, thanks to the explicit training signal related to each word, the discriminator can also be simplified to have a lightweight structure. Compared with the state of the art, our method has a much smaller number of parameters, but still achieves a competitive manipulation performance. Extensive experimental results demonstrate that our method can better disentangle different visual attributes, then correctly map them to corresponding semantic words, and thus achieve a more accurate image modification using natural language descriptions. |
Persistent Identifier | http://hdl.handle.net/10722/308211 |
ISSN | 2020 SCImago Journal Rankings: 1.399 |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Li, B | - |
dc.contributor.author | Qi, X | - |
dc.contributor.author | Torr, P | - |
dc.contributor.author | Lukasiewicz, T | - |
dc.date.accessioned | 2021-11-12T13:44:02Z | - |
dc.date.available | 2021-11-12T13:44:02Z | - |
dc.date.issued | 2020 | - |
dc.identifier.citation | 34th Conference on Neural Information Processing (NeurIPS), Virtual Conference, 6-12 December 2020. In Larochelle, H ... et al (eds.), Advances in Neural Information Processing Systems 33 (NeurIPS 2020) Proceedings | - |
dc.identifier.issn | 1049-5258 | - |
dc.identifier.uri | http://hdl.handle.net/10722/308211 | - |
dc.description.abstract | We propose a novel lightweight generative adversarial network for efficient image manipulation using natural language descriptions. To achieve this, a new word-level discriminator is proposed, which provides the generator with fine-grained training feedback at word-level, to facilitate training a lightweight generator that has a small number of parameters, but can still correctly focus on specific visual attributes of an image, and then edit them without affecting other contents that are not described in the text. Furthermore, thanks to the explicit training signal related to each word, the discriminator can also be simplified to have a lightweight structure. Compared with the state of the art, our method has a much smaller number of parameters, but still achieves a competitive manipulation performance. Extensive experimental results demonstrate that our method can better disentangle different visual attributes, then correctly map them to corresponding semantic words, and thus achieve a more accurate image modification using natural language descriptions. | - |
dc.language | eng | - |
dc.publisher | Neural Information Processing Systems Foundation, Inc. The Journal's web site is located at https://papers.nips.cc/ | - |
dc.relation.ispartof | 34th Conference on Neural Information Processing (NeurIPS) 2020 | - |
dc.relation.ispartof | Advances in Neural Information Processing Systems 33 (NIPS 2020 Proceedings) | - |
dc.title | Lightweight Generative Adversarial Networks for Text-Guided Image Manipulation | - |
dc.type | Conference_Paper | - |
dc.identifier.email | Qi, X: xjqi@eee.hku.hk | - |
dc.identifier.authority | Qi, X=rp02666 | - |
dc.description.nature | published_or_final_version | - |
dc.identifier.hkuros | 329587 | - |
dc.publisher.place | United States | - |