{"url":"/method/specgan","slug":"specgan","name":"SpecGAN","full_name":"SpecGAN","full_name_withheld":false,"description_markdown":"**SpecGAN** is a generative adversarial network method for spectrogram-based, frequency-domain audio generation. The problem is suited for GANs designed for image generation. The model can be approximately inverted. \r\n\r\nTo process audio into suitable spectrograms, the authors perform the short-time Fourier transform with 16 ms windows and 8ms stride, resulting in 128 frequency bins, linearly spaced from 0 to 8 kHz. They take the magnitude of the resultant spectra and scale amplitude values logarithmically to better-align with human perception. They then normalize each frequency bin to have zero mean and unit variance. They clip the spectra to $3$ standard deviations and rescale to $\\left[−1, 1\\right]$.\r\n\r\nThey then use the [DCGAN](https://paperswithcode.com/method/dcgan) approach on the result spectra.","description_state":"present","introduced_year":null,"introduced_by":{"title":"Adversarial Audio Synthesis","paper":"/paper/adversarial-audio-synthesis","first_author":"Chris Donahue","n_authors":3,"url_abs":null,"archive_paper_url":"https://paperswithcode.com/paper/adversarial-audio-synthesis"},"source":{"url":"http://arxiv.org/abs/1802.04208v3","title":"Adversarial Audio Synthesis","url_on_a_paper_host":true},"code_snippet_url":null,"code_snippet_url_on_a_code_host":false,"categories":[{"area":"Audio","area_id":"audio","collection":"Generative Audio Models","url":"/methods/category/generative-audio-models","pwc_aliases":[]}],"n_papers_tagged":1,"archive_num_papers":1,"papers_newest_first":[{"paper":"/paper/adversarial-audio-synthesis","title":"Adversarial Audio Synthesis","date":"2018-02-12","arxiv_id":"1802.04208","n_code_links":22,"syntology":{"ran":2,"of":4,"unverified":2,"pointer_only":4}}],"papers_shown":1,"tasks":[{"task":"/task/audio-generation","name":"Audio Generation","papers":1},{"task":"/task/audio-synthesis","name":"Audio Synthesis","papers":1},{"task":"/task/image-generation","name":"Image Generation","papers":1}],"tasks_shown":3,"n_tasks":3,"usage_by_year":[{"year":"2018","papers":1}],"row_source":"methods_table","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/specgan"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}