Audio-visual scene classification: analysis of DCASE 2021 Challenge submissions
Wang, Shanshan; Heittola, Toni; Mesaros, Annamaria; Virtanen, Tuomas (2021-11-15)
URI
https://dcase.community/documents/workshop2021/proceedings/DCASE2021Workshop_Wang_20.pdfhttps://dcase.community/workshop2021/proceedings
Wang, Shanshan
Heittola, Toni
Mesaros, Annamaria
Virtanen, Tuomas
Teoksen toimittaja(t)
Font, Frederic
Mesaros, Annamaria
P.W. Ellis, Daniel
Fonseca, Eduardo
Fuentes, Magdalena
Elizalde, Benjamin
DCASE
15.11.2021
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202203012247
https://urn.fi/URN:NBN:fi:tuni-202203012247
Kuvaus
Peer reviewed
Tiivistelmä
This paper presents the details of the Audio-Visual Scene Classification task in the DCASE 2021 Challenge (Task 1 Subtask B). The task is concerned with classification using audio and video modalities, using a dataset of synchronized recordings. This task has attracted 43 submissions from 13 different teams around the world. Among all submissions, more than half of the submitted systems have better performance than the baseline. The common techniques among the top systems are the usage of large pretrained models such as ResNet or EfficientNet which are trained for the task-specific problem. Fine-tuning, transfer learning, and data augmentation techniques are also employed to boost the performance. More importantly, multi-modal methods using both audio and video are employed by all the top 5 teams. The best system among all achieved a logloss of 0.195 and accuracy of 93.8\%, compared to the baseline system with logloss of 0.662 and accuracy of 77.1%.
Kokoelmat
- TUNICRIS-julkaisut [18531]