Wassim Bouachir
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Wassim Bouachir has a total of 45 co-authors. Here are the top co-authors by number of shared publications:
- Guillaume-Alexandre Bilodeau
Software Engineering Department Chem. de Poly- technique Polytechnique Montréal 2500 H3T 1J4 Montréal Canada
13 shared publications - Nizar Bouguila
Concordia University, Montreal, Canada
10 shared publications - Alaa Nfissi
Data Science Laboratory, University of Quebec (TELUQ), Montréal, Canada
5 shared publications - Brian L. Mishara
Department of Psychology and Centre for Research and Intervention on Suicide, Ethical Issues and End of Life Practices (CRISE), Université du Québec à Montréal, Montréal, Québec, Canada
5 shared publications - Safwen Naimi
University of Quebec (TÉLUQ) Montreal QC Canada
4 shared publications - Noreen Anwar
LITIV Polytechnique Montréal Montréal Canada
3 shared publications
Latest Publications
- 7 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-8
Parkinson's Disease Severity Assessment from Gait Patterns: A CNN-Transformer Hybrid Approach
Parkinson's disease remains a major medical challenge, with no definitive cure and research focused on symptom management and severity assessment. We propose the CNN-Transformer hybrid approach, which leverages identical spatial and temporal transformer blocks to capture long-range dependencies—key for accurate classification of gait data into four Hoehn and Yahr stages. While prior work confirms transformers' strength in PD severity detection, we optimized the ConvNet block, finding that ConvBiLSTM outperforms alternatives (Inception1D, ConvLSTM) and concurrent methods. The retained ConvBiLSTM model achieves 95.91 % accuracy on the Physionet dataset, surpassing state-of-the-art approaches. This advancement supports more reliable PD progression monitoring and improved patient care.
- 7 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-8
Parkinson's Disease Severity Assessment from Gait Patterns: A CNN-Transformer Hybrid Approach
Parkinson's disease remains a major medical challenge, with no definitive cure and research focused on symptom management and severity assessment. We propose the CNN-Transformer hybrid approach, which leverages identical spatial and temporal transformer blocks to capture long-range dependencies—key for accurate classification of gait data into four Hoehn and Yahr stages. While prior work confirms transformers' strength in PD severity detection, we optimized the ConvNet block, finding that ConvBiLSTM outperforms alternatives (Inception1D, ConvLSTM) and concurrent methods. The retained ConvBiLSTM model achieves 95.91 % accuracy on the Physionet dataset, surpassing state-of-the-art approaches. This advancement supports more reliable PD progression monitoring and improved patient care.
- Open Access
Dual-Stream Attention with Multi-Modal Queries for Object Detection in Transportation Applications
Transformer-based object detectors often struggle with occlusions, fine-grained localization, and computational inefficiency caused by fixed queries and dense attention. We propose DAMM, Dual-stream Attention with Multi-Modal queries, a novel framework introducing both query adaptation and structured cross-attention for improved accuracy and efficiency. DAMM capitalizes on three types of queries: appearance-based queries from vision-language models, positional queries using polygonal embeddings, and random learned queries for general scene coverage. Furthermore, a dual-stream cross-attention module separately refines semantic and spatial features, boosting localization precision in cluttered scenes. We evaluated DAMM on four challenging benchmarks, and it achieved state-of-the-art performance in average precision (AP) and recall, demonstrating the effectiveness of multi-modal query adaptation and dual-stream attention. Source code is at: \href{https://github.com/DET-LIP/DAMM}{GitHub}.
Open AccessDual-Stream Attention with Multi-Modal Queries for Object Detection in Transportation Applications
Transformer-based object detectors often struggle with occlusions, fine-grained localization, and computational inefficiency caused by fixed queries and dense attention. We propose DAMM, Dual-stream Attention with Multi-Modal queries, a novel framework introducing both query adaptation and structured cross-attention for improved accuracy and efficiency. DAMM capitalizes on three types of queries: appearance-based queries from vision-language models, positional queries using polygonal embeddings, and random learned queries for general scene coverage. Furthermore, a dual-stream cross-attention module separately refines semantic and spatial features, boosting localization precision in cluttered scenes. We evaluated DAMM on four challenging benchmarks, and it achieved state-of-the-art performance in average precision (AP) and recall, demonstrating the effectiveness of multi-modal query adaptation and dual-stream attention. Source code is at: \href{https://github.com/DET-LIP/DAMM}{GitHub}.
- 3 June 2025
- Published by Springer Nature in Applied Intelligence
- Vol. 55 (8) , 1-23
From unaltered raw waveform to emotion: Synergizing convolutional and gated recurrent networks for holistic speech emotion analysis
- 3 June 2025
- Published by Springer Nature in Applied Intelligence
- Vol. 55 (8) , 1-23
From unaltered raw waveform to emotion: Synergizing convolutional and gated recurrent networks for holistic speech emotion analysis
- 13 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Affective Computing
- Vol. 16 (3) , 1839-1854
SigWavNet: Learning Multiresolution Signal Wavelet Network for Speech Emotion Recognition
In the field of human-computer interaction and psychological assessment, speech emotion recognition (SER) plays an important role in deciphering emotional states from speech signals. Despite advancements, challenges persist due to system complexity, feature distinctiveness issues, and noise interference. This paper introduces a new end-to-end (E2E) deep learning multi-resolution framework for SER, addressing these limitations by extracting meaningful representations directly from raw waveform speech signals. By leveraging the properties of the fast discrete wavelet transform (FDWT), including the cascade algorithm, conjugate quadrature filter, and coefficient denoising, our approach introduces a learnable model for both wavelet bases and denoising through deep learning techniques. The framework incorporates an activation function for learnable asymmetric hard thresholding of wavelet coefficients. Our approach exploits the capabilities of wavelets for effective localization in both time and frequency domains. We then combine one-dimensional dilated convolutional neural networks (1D dilated CNN) with a spatial attention layer and bidirectional gated recurrent units (Bi-GRU) with a temporal attention layer to efficiently capture the nuanced spatial and temporal characteristics of emotional features. By handling variable-length speech without segmentation and eliminating the need for pre or post-processing, the proposed model outperformed state-of-the-art methods on IEMOCAP and EMO-DB datasets. The source code of this paper is shared on the Github repository: https://github.com/alaaNfissi/SigWavNet-Learning-Multiresolution-Signal-Wavelet-Network-for-Speech-Emotion-Recognition.
- 13 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Affective Computing
- Vol. 16 (3) , 1839-1854
SigWavNet: Learning Multiresolution Signal Wavelet Network for Speech Emotion Recognition
In the field of human-computer interaction and psychological assessment, speech emotion recognition (SER) plays an important role in deciphering emotional states from speech signals. Despite advancements, challenges persist due to system complexity, feature distinctiveness issues, and noise interference. This paper introduces a new end-to-end (E2E) deep learning multi-resolution framework for SER, addressing these limitations by extracting meaningful representations directly from raw waveform speech signals. By leveraging the properties of the fast discrete wavelet transform (FDWT), including the cascade algorithm, conjugate quadrature filter, and coefficient denoising, our approach introduces a learnable model for both wavelet bases and denoising through deep learning techniques. The framework incorporates an activation function for learnable asymmetric hard thresholding of wavelet coefficients. Our approach exploits the capabilities of wavelets for effective localization in both time and frequency domains. We then combine one-dimensional dilated convolutional neural networks (1D dilated CNN) with a spatial attention layer and bidirectional gated recurrent units (Bi-GRU) with a temporal attention layer to efficiently capture the nuanced spatial and temporal characteristics of emotional features. By handling variable-length speech without segmentation and eliminating the need for pre or post-processing, the proposed model outperformed state-of-the-art methods on IEMOCAP and EMO-DB datasets. The source code of this paper is shared on the Github repository: https://github.com/alaaNfissi/SigWavNet-Learning-Multiresolution-Signal-Wavelet-Network-for-Speech-Emotion-Recognition.
- Open Access
- 10 February 2025
- Published by MDPI AG in Journal of Imaging
- Vol. 11 (2) , 52
SAVE: Self-Attention on Visual Embedding for Zero-Shot Generic Object Counting
Zero-shot counting is a subcategory of Generic Visual Object Counting, which aims to count objects from an arbitrary class in a given image. While few-shot counting relies on delivering exemplars to the model to count similar class objects, zero-shot counting automates the operation for faster processing. This paper proposes a fully automated zero-shot method outperforming both zero-shot and few-shot methods. By exploiting feature maps from a pre-trained detection-based backbone, we introduce a new Visual Embedding Module designed to generate semantic embeddings within object contextual information. These embeddings are then fed to a Self-Attention Matching Module to generate an encoded representation for the head counter. Our proposed method has outperformed recent zero-shot approaches, achieving the best Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) results of and , respectively, on the FSC147 dataset. Additionally, our method demonstrates competitive performance compared to few-shot methods, advancing the capabilities of visual object counting in various industrial applications such as tree counting, wildlife animal counting, and medical applications like blood cell counting.- 10 February 2025
- Published by MDPI AG in Journal of Imaging
- Vol. 11 (2) , 52
Open AccessSAVE: Self-Attention on Visual Embedding for Zero-Shot Generic Object Counting
Zero-shot counting is a subcategory of Generic Visual Object Counting, which aims to count objects from an arbitrary class in a given image. While few-shot counting relies on delivering exemplars to the model to count similar class objects, zero-shot counting automates the operation for faster processing. This paper proposes a fully automated zero-shot method outperforming both zero-shot and few-shot methods. By exploiting feature maps from a pre-trained detection-based backbone, we introduce a new Visual Embedding Module designed to generate semantic embeddings within object contextual information. These embeddings are then fed to a Self-Attention Matching Module to generate an encoded representation for the head counter. Our proposed method has outperformed recent zero-shot approaches, achieving the best Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) results of and , respectively, on the FSC147 dataset. Additionally, our method demonstrates competitive performance compared to few-shot methods, advancing the capabilities of visual object counting in various industrial applications such as tree counting, wildlife animal counting, and medical applications like blood cell counting. - 19 September 2024
- Published by Springer Nature
- p. 256-268
Deep Multiresolution Wavelet Transform for Speech Emotion Assessment of High-Risk Suicide Callers
- 19 September 2024
- Published by Springer Nature
- p. 256-268
Deep Multiresolution Wavelet Transform for Speech Emotion Assessment of High-Risk Suicide Callers
- 31 May 2024
- Published by Springer Nature in Applied Intelligence
- Vol. 54 (11-12) , 7046-7069
Unveiling hidden factors: explainable AI for feature boosting in speech emotion recognition
- 31 May 2024
- Published by Springer Nature in Applied Intelligence
- Vol. 54 (11-12) , 7046-7069
Unveiling hidden factors: explainable AI for feature boosting in speech emotion recognition
- Open Access
STF: Spatio-Temporal Fusion Module for Improving Video Object Detection
Consecutive frames in a video contain redundancy, but they may also contain relevant complementary information for the detection task. The objective of our work is to leverage this complementary information to improve detection. Therefore, we propose a spatio-temporal fusion framework (STF). We first introduce multi-frame and single-frame attention modules that allow a neural network to share feature maps between nearby frames to obtain more robust object representations. Second, we introduce a dual-frame fusion module that merges feature maps in a learnable manner to improve them. Our evaluation is conducted on three different benchmarks including video sequences of moving road users. The performed experiments demonstrate that the proposed spatio-temporal fusion module leads to improved detection performance compared to baseline object detectors. Code is available at https://github.com/noreenanwar/STF-module
Open AccessSTF: Spatio-Temporal Fusion Module for Improving Video Object Detection
Consecutive frames in a video contain redundancy, but they may also contain relevant complementary information for the detection task. The objective of our work is to leverage this complementary information to improve detection. Therefore, we propose a spatio-temporal fusion framework (STF). We first introduce multi-frame and single-frame attention modules that allow a neural network to share feature maps between nearby frames to obtain more robust object representations. Second, we introduce a dual-frame fusion module that merges feature maps in a learnable manner to improve them. Our evaluation is conducted on three different benchmarks including video sequences of moving road users. The performed experiments demonstrate that the proposed spatio-temporal fusion module leads to improved detection performance compared to baseline object detectors. Code is available at https://github.com/noreenanwar/STF-module
- Open Access
- 15 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 543-549
Iterative Feature Boosting for Explainable Speech Emotion Recognition
In speech emotion recognition (SER), using predefined features without considering their practical importance may lead to high dimensional datasets, including redundant and irrelevant information. Consequently, high-dimensional learning often results in decreasing model accuracy while increasing computational complexity. Our work underlines the importance of carefully considering and analyzing features in order to build efficient SER systems. We present a new supervised SER method based on an efficient feature engineering approach. We pay particular attention to the explainability of results to evaluate feature relevance and refine feature sets. This is performed iteratively through feature evaluation loop, using Shapley values to boost feature selection and improve overall framework performance. Our approach allows thus to balance the benefits between model performance and transparency. The proposed method outperforms human-level performance (HLP) and state-of-the-art machine learning methods in emotion recognition on the TESS dataset. The source code of this paper is publicly available at Iterative-Feature-Boosting-for-Explainable-Speech-Emotion-Recoanition.
- 15 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 543-549
Open AccessIterative Feature Boosting for Explainable Speech Emotion Recognition
In speech emotion recognition (SER), using predefined features without considering their practical importance may lead to high dimensional datasets, including redundant and irrelevant information. Consequently, high-dimensional learning often results in decreasing model accuracy while increasing computational complexity. Our work underlines the importance of carefully considering and analyzing features in order to build efficient SER systems. We present a new supervised SER method based on an efficient feature engineering approach. We pay particular attention to the explainability of results to evaluate feature relevance and refine feature sets. This is performed iteratively through feature evaluation loop, using Shapley values to boost feature selection and improve overall framework performance. Our approach allows thus to balance the benefits between model performance and transparency. The proposed method outperforms human-level performance (HLP) and state-of-the-art machine learning methods in emotion recognition on the TESS dataset. The source code of this paper is publicly available at Iterative-Feature-Boosting-for-Explainable-Speech-Emotion-Recoanition.
- Open Access
- 15 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 814-819
HCT: Hybrid Convnet-Transformer for Parkinson's Disease Detection and Severity Prediction from Gait
In this paper, we propose a novel deep learning method based on a new Hybrid ConvNet-Transformer archi-tecture to detect and stage Parkinson's disease (PD) from gait data. We adopt a two-step approach by dividing the problem into two sub-problems. Our Hybrid ConvNet-Transformer model first distinguishes healthy versus parkinsonian patients. If the patient is parkinsonian, a multi-class Hybrid ConvNet-Transformer model determines the Hoehn and Yahr (H&Y) score to assess the PD severity stage. Our hybrid architecture exploits the strengths of both Convolutional Neural Networks (ConvNets) and Transformers to accurately detect PD and determine the severity stage. In particular, we take advantage of ConvNets to capture local patterns and correlations in the data, while we exploit Transformers for handling long-term dependencies in the input signal. We show that our hybrid method achieves superior performance when compared to other state-of-the-art methods, with a PD detection accuracy of 97% and a severity staging accuracy of 87%. Our source code is available at https://github.com/SafwenNaimi.
- 15 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 814-819
Open AccessHCT: Hybrid Convnet-Transformer for Parkinson's Disease Detection and Severity Prediction from Gait
In this paper, we propose a novel deep learning method based on a new Hybrid ConvNet-Transformer archi-tecture to detect and stage Parkinson's disease (PD) from gait data. We adopt a two-step approach by dividing the problem into two sub-problems. Our Hybrid ConvNet-Transformer model first distinguishes healthy versus parkinsonian patients. If the patient is parkinsonian, a multi-class Hybrid ConvNet-Transformer model determines the Hoehn and Yahr (H&Y) score to assess the PD severity stage. Our hybrid architecture exploits the strengths of both Convolutional Neural Networks (ConvNets) and Transformers to accurately detect PD and determine the severity stage. In particular, we take advantage of ConvNets to capture local patterns and correlations in the data, while we exploit Transformers for handling long-term dependencies in the input signal. We show that our hybrid method achieves superior performance when compared to other state-of-the-art methods, with a PD detection accuracy of 97% and a severity staging accuracy of 87%. Our source code is available at https://github.com/SafwenNaimi.