Samuele Cornell
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Samuele Cornell has a total of 46 co-authors. Here are the top co-authors by number of shared publications:
- Stefano Squartini
Department of Information Engineering, Università Politecnica delle Marche, Via Brecce Bianche 12, 60131 Ancona, Italy
7 shared publications - Alessio Brutti
Fondazione Bruno Kessler Italy
3 shared publications - Carlo Aironi
Universit Politecnica delle Marche, Ancona, Italy
3 shared publications Stefano Squartini
3 shared publicationsEnrico Zovato
2 shared publicationsGiovanni Morrone
2 shared publications
Latest Publications
- 1 November 2025
- Published by Elsevier in Computer Speech & Language
- Vol. 97
Recent trends in distant conversational speech recognition: A review of CHiME-7 and 8 DASR challenges
- 1 November 2025
- Published by Elsevier in Computer Speech & Language
- Vol. 97
Recent trends in distant conversational speech recognition: A review of CHiME-7 and 8 DASR challenges
- Open Access
- 11 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 33 (23299290) , 199-210
Complex-bin2bin: A Latency-Flexible Generative Neural Model for Audio Packet Loss Concealment
Despite the significant advancements in networking technologies, transmission of data packets in real-time, particularly in speech communications, continues to face challenges due to the possibility of data loss. This loss not only compromises sound quality but also diminishes overall intelligibility. In such cases, Packet Loss Concealment (PLC) techniques could help by reconstructing the missing content and restoring the audio quality. This work proposes a novel method, that improves previous time-frequency generative inpainting approaches. Compared to other state-of-the-art methods, our proposed approach has the flexibility to restore lost packets either in real-time at low latency or in offline mode, without the need to retrain the network. Evaluations conducted against a recent state-of-the-art method, ranked at the top of the 2022 Microsoft PLC competition, and against four DNN-based PLC solutions from the literature, show superior scores in terms of task-specific metrics. The method has also been tested in more challenging scenarios than aforementioned ones, with packet loss rates of up to 50%, showing the ability to help automatic speech recognition (ASR) systems reduce word error rate (WER) by up to almost 50% relative improvement. Additionally, a comparative subjective evaluation has been conducted, confirming the effectiveness of the proposed method in relation to the state of the art. The code is made available in the project repository https://github.com/aircarlo/cplx_bin2bin .
- 11 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech and Language Processing
- Vol. 33 (23299290) , 199-210
Open AccessComplex-bin2bin: A Latency-Flexible Generative Neural Model for Audio Packet Loss Concealment
Despite the significant advancements in networking technologies, transmission of data packets in real-time, particularly in speech communications, continues to face challenges due to the possibility of data loss. This loss not only compromises sound quality but also diminishes overall intelligibility. In such cases, Packet Loss Concealment (PLC) techniques could help by reconstructing the missing content and restoring the audio quality. This work proposes a novel method, that improves previous time-frequency generative inpainting approaches. Compared to other state-of-the-art methods, our proposed approach has the flexibility to restore lost packets either in real-time at low latency or in offline mode, without the need to retrain the network. Evaluations conducted against a recent state-of-the-art method, ranked at the top of the 2022 Microsoft PLC competition, and against four DNN-based PLC solutions from the literature, show superior scores in terms of task-specific metrics. The method has also been tested in more challenging scenarios than aforementioned ones, with packet loss rates of up to 50%, showing the ability to help automatic speech recognition (ASR) systems reduce word error rate (WER) by up to almost 50% relative improvement. Additionally, a comparative subjective evaluation has been conducted, confirming the effectiveness of the proposed method in relation to the state of the art. The code is made available in the project repository https://github.com/aircarlo/cplx_bin2bin .
- 17 January 2024
- Published by World Scientific Pub Co Pte Ltd in International Journal of Neural Systems
- Vol. 34 (03) , 2450011
A Graph-Based Neural Approach to Linear Sum Assignment Problems
Linear assignment problems are well-known combinatorial optimization problems involving domains such as logistics, robotics and telecommunications. In general, obtaining an optimal solution to such problems is computationally infeasible even in small settings, so heuristic algorithms are often used to find near-optimal solutions. In order to attain the right assignment permutation, this study investigates a general-purpose learning strategy that uses a bipartite graph to describe the problem structure and a message passing Graph Neural Network (GNN) model to learn the correct mapping. Comparing the proposed structure with two existing DNN solutions, simulation results show that the proposed approach significantly improves classification accuracy, proving to be very efficient in terms of processing time and memory requirements, due to its inherent parameter sharing capability. Among the many practical uses that require solving allocation problems in everyday scenarios, we decided to apply the proposed approach to address the scheduling of electric smart meters access within an electricity distribution smart grid infrastructure, since near-real-time energy monitoring is a key element of the green transition that has become increasingly important in recent times. The results obtained show that the proposed graph-based solver, although sub-optimal, exhibits the highest scalability, compared with other state-of-the-art heuristic approaches. To foster the reproducibility of the results, we made the code available at https://github.com/aircarlo/GNN_LSAP .
- 17 January 2024
- Published by World Scientific Pub Co Pte Ltd in International Journal of Neural Systems
- Vol. 34 (03) , 2450011
A Graph-Based Neural Approach to Linear Sum Assignment Problems
Linear assignment problems are well-known combinatorial optimization problems involving domains such as logistics, robotics and telecommunications. In general, obtaining an optimal solution to such problems is computationally infeasible even in small settings, so heuristic algorithms are often used to find near-optimal solutions. In order to attain the right assignment permutation, this study investigates a general-purpose learning strategy that uses a bipartite graph to describe the problem structure and a message passing Graph Neural Network (GNN) model to learn the correct mapping. Comparing the proposed structure with two existing DNN solutions, simulation results show that the proposed approach significantly improves classification accuracy, proving to be very efficient in terms of processing time and memory requirements, due to its inherent parameter sharing capability. Among the many practical uses that require solving allocation problems in everyday scenarios, we decided to apply the proposed approach to address the scheduling of electric smart meters access within an electricity distribution smart grid infrastructure, since near-real-time energy monitoring is a key element of the green transition that has become increasingly important in recent times. The results obtained show that the proposed graph-based solver, although sub-optimal, exhibits the highest scalability, compared with other state-of-the-art heuristic approaches. To foster the reproducibility of the results, we made the code available at https://github.com/aircarlo/GNN_LSAP .
- Open Access
- 20 November 2023
- Published by The Open Journal in Journal of Open Source Software
- Vol. 8 (91) , 5403
Software Design and User Interface of ESPnet-SE++: Speech Enhancement for Robust Speech Processing
- 20 November 2023
- Published by The Open Journal in Journal of Open Source Software
- Vol. 8 (91) , 5403
Open AccessSoftware Design and User Interface of ESPnet-SE++: Speech Enhancement for Robust Speech Processing
- Open Access
A Time-Frequency Generative Adversarial based method for Audio Packet Loss Concealment
Packet loss is a major cause of voice quality degradation in VoIP transmissions with serious impact on intelligibility and user experience. This paper describes a system based on a generative adversarial approach, which aims to repair the lost fragments during the transmission of audio streams. Inspired by the powerful image-to-image translation capability of Generative Adversarial Networks (GANs), we propose bin2bin, an improved pix2pix framework to achieve the translation task from magnitude spectrograms of audio frames with lost packets, to noncorrupted speech spectrograms. In order to better maintain the structural information after spectrogram translation, this paper introduces the combination of two STFT-based loss functions, mixed with the traditional GAN objective. Furthermore, we employ a modified PatchGAN structure as discriminator and we lower the concealment time by a proper initialization of the phase reconstruction algorithm. Experimental results show that the proposed method has obvious advantages when compared with the current state-of-the-art methods, as it can better handle both high packet loss rates and large gaps.
Open AccessA Time-Frequency Generative Adversarial based method for Audio Packet Loss Concealment
Packet loss is a major cause of voice quality degradation in VoIP transmissions with serious impact on intelligibility and user experience. This paper describes a system based on a generative adversarial approach, which aims to repair the lost fragments during the transmission of audio streams. Inspired by the powerful image-to-image translation capability of Generative Adversarial Networks (GANs), we propose bin2bin, an improved pix2pix framework to achieve the translation task from magnitude spectrograms of audio frames with lost packets, to noncorrupted speech spectrograms. In order to better maintain the structural information after spectrogram translation, this paper introduces the combination of two STFT-based loss functions, mixed with the traditional GAN objective. Furthermore, we employ a modified PatchGAN structure as discriminator and we lower the concealment time by a proper initialization of the phase reconstruction algorithm. Experimental results show that the proposed method has obvious advantages when compared with the current state-of-the-art methods, as it can better handle both high packet loss rates and large gaps.
- Open Access
- 29 May 2023
- Published by Elsevier in Computer Speech & Language
- Vol. 82
An experimental review of speaker diarization methods with application to two-speaker conversational telephone speech recordings
- 29 May 2023
- Published by Elsevier in Computer Speech & Language
- Vol. 82
Open AccessAn experimental review of speaker diarization methods with application to two-speaker conversational telephone speech recordings
- Open Access
An Experimental Review of Speaker Diarization methods with application to Two-Speaker Conversational Telephone Speech recordings
We performed an experimental review of current diarization systems for the conversational telephone speech (CTS) domain. In detail, we considered a total of eight different algorithms belonging to clustering-based, end-to-end neural diarization (EEND), and speech separation guided diarization (SSGD) paradigms. We studied the inference-time computational requirements and diarization accuracy on four CTS datasets with different characteristics and languages. We found that, among all methods considered, EEND-vector clustering (EEND-VC) offers the best trade-off in terms of computing requirements and performance. More in general, EEND models have been found to be lighter and faster in inference compared to clustering-based methods. However, they also require a large amount of diarization-oriented annotated data. In particular EEND-VC performance in our experiments degraded when the dataset size was reduced, whereas self-attentive EEND (SA-EEND) was less affected. We also found that SA-EEND gives less consistent results among all the datasets compared to EEND-VC, with its performance degrading on long conversations with high speech sparsity. Clustering-based diarization systems, and in particular VBx, instead have more consistent performance compared to SA-EEND but are outperformed by EEND-VC. The gap with respect to this latter is reduced when overlap-aware clustering methods are considered. SSGD is the most computationally demanding method, but it could be convenient if speech recognition has to be performed. Its performance is close to SA-EEND but degrades significantly when the training and inference data characteristics are less matched.
Open AccessAn Experimental Review of Speaker Diarization methods with application to Two-Speaker Conversational Telephone Speech recordings
We performed an experimental review of current diarization systems for the conversational telephone speech (CTS) domain. In detail, we considered a total of eight different algorithms belonging to clustering-based, end-to-end neural diarization (EEND), and speech separation guided diarization (SSGD) paradigms. We studied the inference-time computational requirements and diarization accuracy on four CTS datasets with different characteristics and languages. We found that, among all methods considered, EEND-vector clustering (EEND-VC) offers the best trade-off in terms of computing requirements and performance. More in general, EEND models have been found to be lighter and faster in inference compared to clustering-based methods. However, they also require a large amount of diarization-oriented annotated data. In particular EEND-VC performance in our experiments degraded when the dataset size was reduced, whereas self-attentive EEND (SA-EEND) was less affected. We also found that SA-EEND gives less consistent results among all the datasets compared to EEND-VC, with its performance degrading on long conversations with high speech sparsity. Clustering-based diarization systems, and in particular VBx, instead have more consistent performance compared to SA-EEND but are outperformed by EEND-VC. The gap with respect to this latter is reduced when overlap-aware clustering methods are considered. SSGD is the most computationally demanding method, but it could be convenient if speech recognition has to be performed. Its performance is close to SA-EEND but degrades significantly when the training and inference data characteristics are less matched.
- Open Access
- 9 January 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 260-265
End-to-End Integration of Speech Recognition, Dereverberation, Beamforming, and Self-Supervised Learning Representation
Self-supervised learning representation (SSLR) has demonstrated its significant effectiveness in automatic speech recognition (ASR), mainly with clean speech. Recent work pointed out the strength of integrating SSLR with single-channel speech enhancement for ASR in noisy environments. This paper further advances this integration by dealing with multi-channel input. We propose a novel end-to-end architecture by integrating dereverberation, beamforming, SSLR, and ASR within a single neural network. Our system achieves the best performance reported in the literature on the CHiME-4 6-channel track with a word error rate (WER) of 1.77%. While the WavLM-based strong SSLR demonstrates promising results by itself, the end-to-end integration with the weighted power minimization distortionless response beamformer, which simultaneously performs dereverberation and denoising, improves WER significantly. Its effectiveness is also validated on the REVERB dataset.
- 9 January 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 260-265
Open AccessEnd-to-End Integration of Speech Recognition, Dereverberation, Beamforming, and Self-Supervised Learning Representation
Self-supervised learning representation (SSLR) has demonstrated its significant effectiveness in automatic speech recognition (ASR), mainly with clean speech. Recent work pointed out the strength of integrating SSLR with single-channel speech enhancement for ASR in noisy environments. This paper further advances this integration by dealing with multi-channel input. We propose a novel end-to-end architecture by integrating dereverberation, beamforming, SSLR, and ASR within a single neural network. Our system achieves the best performance reported in the literature on the CHiME-4 6-channel track with a word error rate (WER) of 1.77%. While the WavLM-based strong SSLR demonstrates promising results by itself, the end-to-end integration with the weighted power minimization distortionless response beamformer, which simultaneously performs dereverberation and denoising, improves WER significantly. Its effectiveness is also validated on the REVERB dataset.
- Open Access
Low-Latency Speech Separation Guided Diarization for Telephone Conversations
In this paper, we carry out an analysis on the use of speech separation guided diarization (SSGD) in telephone conversations. SSGD performs diarization by separating the speakers signals and then applying voice activity detection on each estimated speaker signal. In particular, we compare two low-latency speech separation models. Moreover, we show a post-processing algorithm that significantly reduces the false alarm errors of a SSGD pipeline. We perform our experiments on two datasets: Fisher Corpus Part 1 and CALLHOME, evaluating both separation and diarization metrics. Notably, our SSGD DPRNN-based online model achieves 11.1% DER on CALLHOME, comparable with most state-of-the-art end-to-end neural diarization models despite being trained on an order of magnitude less data and having considerably lower latency, i.e., 0.1 vs. 10 seconds. We also show that the separated signals can be readily fed to a speech recognition back-end with performance close to the oracle source signals.
Open AccessLow-Latency Speech Separation Guided Diarization for Telephone Conversations
In this paper, we carry out an analysis on the use of speech separation guided diarization (SSGD) in telephone conversations. SSGD performs diarization by separating the speakers signals and then applying voice activity detection on each estimated speaker signal. In particular, we compare two low-latency speech separation models. Moreover, we show a post-processing algorithm that significantly reduces the false alarm errors of a SSGD pipeline. We perform our experiments on two datasets: Fisher Corpus Part 1 and CALLHOME, evaluating both separation and diarization metrics. Notably, our SSGD DPRNN-based online model achieves 11.1% DER on CALLHOME, comparable with most state-of-the-art end-to-end neural diarization models despite being trained on an order of magnitude less data and having considerably lower latency, i.e., 0.1 vs. 10 seconds. We also show that the separated signals can be readily fed to a speech recognition back-end with performance close to the oracle source signals.
- 1 March 2022
- Published by Elsevier in Computer Speech & Language
- Vol. 72
Overlapped Speech Detection and speaker counting using distant microphone arrays
- 1 March 2022
- Published by Elsevier in Computer Speech & Language
- Vol. 72
Overlapped Speech Detection and speaker counting using distant microphone arrays