2 code implementations • 9 Apr 2024 • Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Noboru Harada, Kunio Kashino
This study proposes Masked Modeling Duo (M2D), an improved masked prediction SSL, which learns by predicting representations of masked input signals that serve as training signals.
1 code implementation • 13 Sep 2023 • Shinnosuke Matsuo, Xiaomeng Wu, Gantugs Atarsaikhan, Akisato Kimura, Kunio Kashino, Brian Kenji Iwana, Seiichi Uchida
Unlike other learnable models using DTW for warping, our model predicts all local correspondences between two time series and is trained based on metric learning, which enables it to learn the optimal data-dependent warping for the target task.
1 code implementation • 23 Aug 2023 • Daiki Takeuchi, Yasunori Ohishi, Daisuke Niizumi, Noboru Harada, Kunio Kashino
We proposed Audio Difference Captioning (ADC) as a new extension task of audio captioning for describing the semantic differences between input pairs of similar but slightly different audio clips.
1 code implementation • 23 May 2023 • Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Noboru Harada, Kunio Kashino
Self-supervised learning general-purpose audio representations have demonstrated high performance in a variety of tasks.
1 code implementation • 26 Oct 2022 • Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Noboru Harada, Kunio Kashino
We propose a new method, Masked Modeling Duo (M2D), that learns representations directly while obtaining training signals using only masked patches.
Ranked #1 on Speaker Identification on VoxCeleb1 (using extra training data)
1 code implementation • 14 Sep 2022 • Xiaomeng Wu, Takahito Kawanishi, Kunio Kashino
Existing image enhancement methods fall short of expectations because with them it is difficult to improve global and local image contrast simultaneously.
1 code implementation • 14 Sep 2022 • Xiaomeng Wu, Yongqing Sun, Akisato Kimura, Kunio Kashino
Despite recent advances in image enhancement, it remains difficult for existing approaches to adaptively improve the brightness and contrast for both low-light and normal-light images.
no code implementations • 25 Jul 2022 • Yasunori Ohishi, Marc Delcroix, Tsubasa Ochiai, Shoko Araki, Daiki Takeuchi, Daisuke Niizumi, Akisato Kimura, Noboru Harada, Kunio Kashino
We use it to bridge modality-dependent information, i. e., the speech segments in the mixture, and the specified, modality-independent concept.
1 code implementation • 20 Jul 2022 • Daiki Takeuchi, Yasunori Ohishi, Daisuke Niizumi, Noboru Harada, Kunio Kashino
While the range of conventional content-based audio retrieval is limited to audio that is similar to the query audio, the proposed method can adjust the retrieval range by adding an embedding of the auxiliary text query-modifier to the embedding of the query sample audio in a shared latent space.
1 code implementation • 17 May 2022 • Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Noboru Harada, Kunio Kashino
This approach improves the utility of frequency and channel information in downstream processes, and combines the effectiveness of middle and late layer features for different tasks.
1 code implementation • 26 Apr 2022 • Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Noboru Harada, Kunio Kashino
In this paper, we seek to learn audio representations from the input itself as supervision using a pretext task of auto-encoding of masked spectrogram patches, Masked Spectrogram Modeling (MSM, a variant of Masked Image Modeling applied to audio spectrogram).
1 code implementation • 15 Apr 2022 • Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Noboru Harada, Kunio Kashino
In this study, we hypothesize that representations effective for general audio tasks should provide multiple aspects of robust features of the input sound.
1 code implementation • 28 Mar 2021 • Shinnosuke Matsuo, Xiaomeng Wu, Gantugs Atarsaikhan, Akisato Kimura, Kunio Kashino, Brian Kenji Iwana, Seiichi Uchida
This approach adapts a parameterized attention model to time warping for greater and more adaptive temporal invariance.
2 code implementations • 11 Mar 2021 • Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Noboru Harada, Kunio Kashino
Inspired by the recent progress in self-supervised learning for computer vision that generates supervision using data augmentations, we explore a new general-purpose audio representation learning approach.
no code implementations • 24 Sep 2020 • Daiki Takeuchi, Yuma Koizumi, Yasunori Ohishi, Noboru Harada, Kunio Kashino
The system we used for Task 6 (Automated Audio Captioning)of the Detection and Classification of Acoustic Scenes and Events(DCASE) 2020 Challenge combines three elements, namely, dataaugmentation, multi-task learning, and post-processing, for audiocaptioning.
no code implementations • 1 Jul 2020 • Yuma Koizumi, Daiki Takeuchi, Yasunori Ohishi, Noboru Harada, Kunio Kashino
This technical report describes the system participating to the Detection and Classification of Acoustic Scenes and Events (DCASE) 2020 Challenge, Task 6: automated audio captioning.
Ranked #4 on Audio captioning on Clotho
no code implementations • CVPR 2018 • Takuhiro Kaneko, Kaoru Hiramatsu, Kunio Kashino
This paper proposes the decision tree latent controller generative adversarial network (DTLC-GAN), an extension of a GAN that can learn hierarchically interpretable representations without relying on detailed supervision.
no code implementations • 18 May 2018 • Chihiro Watanabe, Kaoru Hiramatsu, Kunio Kashino
Interpretability has become an important issue in the machine learning field, along with the success of layered neural networks in various practical tasks.
no code implementations • 13 Apr 2018 • Chihiro Watanabe, Kaoru Hiramatsu, Kunio Kashino
We show experimentally that our proposed method can reveal the role of each part of a layered neural network by applying the neural networks to three types of data sets, extracting communities from the trained network, and applying the proposed method to the community structure.
no code implementations • CVPR 2017 • Takuhiro Kaneko, Kaoru Hiramatsu, Kunio Kashino
This controller is based on a novel generative model called the conditional filtered generative adversarial network (CFGAN), which is an extension of the conventional conditional GAN (CGAN) that incorporates a filtering architecture into the generator input.
no code implementations • 1 Mar 2017 • Chihiro Watanabe, Kaoru Hiramatsu, Kunio Kashino
And (3) data analysis: in practical data it reveals the community structure in the input, hidden, and output layers, which serves as a clue for discovering knowledge from a trained neural network.
no code implementations • ICCV 2015 • Xiaomeng Wu, Kunio Kashino
Hough voting in a geometric transformation space allows us to realize spatial verification, but remains sensitive to feature detection errors because of the inflexible quantization of single feature correspondences.