Zobrazeno 1 - 10
of 21
pro vyhledávání: '"Primus, Paul"'
Autor:
Schmid, Florian, Morocutti, Tobias, Foscarin, Francesco, Schlüter, Jan, Primus, Paul, Widmer, Gerhard
We propose a pre-training pipeline for audio spectrogram transformers for frame-level sound event detection tasks. On top of common pre-training steps, we add a meticulously designed training routine on AudioSet frame-level annotations. This includes
Externí odkaz:
http://arxiv.org/abs/2409.09546
Dual-encoder-based audio retrieval systems are commonly optimized with contrastive learning on a set of matching and mismatching audio-caption pairs. This leads to a shared embedding space in which corresponding items from the two modalities end up c
Externí odkaz:
http://arxiv.org/abs/2408.11641
Query-by-Vocal Imitation (QBV) is about searching audio files within databases using vocal imitations created by the user's voice. Since most humans can effectively communicate sound concepts through voice, QBV offers the more intuitive and convenien
Externí odkaz:
http://arxiv.org/abs/2408.11638
This technical report describes the CP-JKU team's submission for Task 4 Sound Event Detection with Heterogeneous Training Datasets and Potentially Missing Labels of the DCASE 24 Challenge. We fine-tune three large Audio Spectrogram Transformers, PaSS
Externí odkaz:
http://arxiv.org/abs/2408.00791
A central problem in building effective sound event detection systems is the lack of high-quality, strongly annotated sound event datasets. For this reason, Task 4 of the DCASE 2024 challenge proposes learning from two heterogeneous datasets, includi
Externí odkaz:
http://arxiv.org/abs/2407.12997
Autor:
Primus, Paul, Widmer, Gerhard
Matching raw audio signals with textual descriptions requires understanding the audio's content and the description's semantics and then drawing connections between the two modalities. This paper investigates a hybrid retrieval system that utilizes a
Externí odkaz:
http://arxiv.org/abs/2406.15897
Autor:
Schmid, Florian, Primus, Paul, Heittola, Toni, Mesaros, Annamaria, Martín-Morató, Irene, Koutini, Khaled, Widmer, Gerhard
This article describes the Data-Efficient Low-Complexity Acoustic Scene Classification Task in the DCASE 2024 Challenge and the corresponding baseline system. The task setup is a continuation of previous editions (2022 and 2023), which focused on rec
Externí odkaz:
http://arxiv.org/abs/2405.10018
This work presents a text-to-audio-retrieval system based on pre-trained text and spectrogram transformers. Our method projects recordings and textual descriptions into a shared audio-caption space in which related examples from different modalities
Externí odkaz:
http://arxiv.org/abs/2308.04258
Autor:
Primus, Paul, Widmer, Gerhard
The absence of large labeled datasets remains a significant challenge in many application areas of deep learning. Researchers and practitioners typically resort to transfer learning and data augmentation to alleviate this issue. We study these strate
Externí odkaz:
http://arxiv.org/abs/2208.11460
Autor:
Primus, Paul, Widmer, Gerhard
Standard machine learning models for tagging and classifying acoustic signals cannot handle classes that were not seen during training. Zero-Shot (ZS) learning overcomes this restriction by predicting classes based on adaptable class descriptions. Th
Externí odkaz:
http://arxiv.org/abs/2208.11402