Audio retrieval with natural language queries
We consider the task of retrieving audio using free-form natural language queries. To study this problem, which has received limited attention in the existing literature, we introduce challenging new benchmarks for text-based audio retrieval using text annotations sourced from the AudioCaps and Clot...
| Main Authors: | , , , , |
|---|---|
| Format: | Conference item |
| Language: | English |
| Published: |
International Speech Communication Association
2021
|
| _version_ | 1826307889074012160 |
|---|---|
| author | Oncescu, A-M Sophia, AS Henriques, JF Akata, Z Albanie, S |
| author_facet | Oncescu, A-M Sophia, AS Henriques, JF Akata, Z Albanie, S |
| author_sort | Oncescu, A-M |
| collection | OXFORD |
| description | We consider the task of retrieving audio using free-form natural language queries. To study this problem, which has received limited attention in the existing literature, we introduce challenging new benchmarks for text-based audio retrieval using text annotations sourced from the AudioCaps and Clotho datasets. We then employ these benchmarks to establish baselines for cross-modal audio retrieval, where we demonstrate the benefits of pre-training on diverse audio tasks. We hope that our benchmarks will inspire further research into cross-modal text-based audio retrieval with free-form text queries. |
| first_indexed | 2024-03-07T07:09:47Z |
| format | Conference item |
| id | oxford-uuid:cc107781-c321-459d-ba0b-70cef6c52222 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T07:09:47Z |
| publishDate | 2021 |
| publisher | International Speech Communication Association |
| record_format | dspace |
| spelling | oxford-uuid:cc107781-c321-459d-ba0b-70cef6c522222022-06-13T12:45:52ZAudio retrieval with natural language queriesConference itemhttp://purl.org/coar/resource_type/c_5794uuid:cc107781-c321-459d-ba0b-70cef6c52222EnglishSymplectic ElementsInternational Speech Communication Association2021Oncescu, A-MSophia, ASHenriques, JFAkata, ZAlbanie, SWe consider the task of retrieving audio using free-form natural language queries. To study this problem, which has received limited attention in the existing literature, we introduce challenging new benchmarks for text-based audio retrieval using text annotations sourced from the AudioCaps and Clotho datasets. We then employ these benchmarks to establish baselines for cross-modal audio retrieval, where we demonstrate the benefits of pre-training on diverse audio tasks. We hope that our benchmarks will inspire further research into cross-modal text-based audio retrieval with free-form text queries. |
| spellingShingle | Oncescu, A-M Sophia, AS Henriques, JF Akata, Z Albanie, S Audio retrieval with natural language queries |
| title | Audio retrieval with natural language queries |
| title_full | Audio retrieval with natural language queries |
| title_fullStr | Audio retrieval with natural language queries |
| title_full_unstemmed | Audio retrieval with natural language queries |
| title_short | Audio retrieval with natural language queries |
| title_sort | audio retrieval with natural language queries |
| work_keys_str_mv | AT oncescuam audioretrievalwithnaturallanguagequeries AT sophiaas audioretrievalwithnaturallanguagequeries AT henriquesjf audioretrievalwithnaturallanguagequeries AT akataz audioretrievalwithnaturallanguagequeries AT albanies audioretrievalwithnaturallanguagequeries |