Mohammad Ali Akhaee
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Mohammad Ali Akhaee has a total of 85 co-authors. Here are the top co-authors by number of shared publications:
- Ramin Toosi
University of Tehran
15 shared publications - Mohammad-Reza Abolghasemi Dehaqani
School of Electrical and Computer Engineering, College of Engineering, University of Tehran, Tehran, Iran
4 shared publications - Ali Ameri
Department of Biomedical Engineering and Biophysics, School of Medicine, Shahid Beheshti University of Medical Sciences, Tehran, Iran
3 shared publications Jalaledin Noroozi
School of Cognitive Sciences, Institute for Research in Fundamental Sciences
3 shared publicationsRoxana Koushki
School of Cognitive Sciences, Institute for Research in Fundamental Sciences
3 shared publicationsAbdol-Hossein Vahabie
School of Cognitive Sciences, Institute for Research in Fundamental Sciences (IPM)
2 shared publications
Latest Publications
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 26  (19) , 29595-29607
Deep Multi-Task Locomotion-Invariant Gesture Recognition With FMCW Radar
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 26  (19) , 29595-29607
Deep Multi-Task Locomotion-Invariant Gesture Recognition With FMCW Radar
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 26  (20) , 1
Deep Gesture Recognition under Data Loss
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 26  (20) , 1
Deep Gesture Recognition under Data Loss
- Open Access
- 21 November 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15  (1) , 41347
A multi-head YOLOv12 with self-supervised pretraining for urinary sediment particle detection
Automatic and reliable urine sediment analysis is essential for timely diagnosis and management of renal and urinary disorders. However, manual methods are time-consuming, subjective, and limited by operator abilities. In this study, we propose a novel deep learning method based on a multi-head YOLOv12 architecture combined with self-supervised pretraining and advanced inference through Slicing Aided Hyper Inference (SAHI) to effectively address these challenges. Unlike prior methods that employed a single detection head, our architecture features six specialized and independent detection heads: Cells, Casts, Crystals, Microorganisms/Yeast, Artifact, and Others, enabling simultaneous and fine-grained classification of the full spectrum of urine sediment particles, including all relevant subclasses. To facilitate robust training, we created a large-scale dataset (OpenUrine) encompassing 790 labeled images with over 31,285 bounding boxes across 39 categories, and 5640 unlabeled images for self-supervised learning. Evaluated on this complex 39-class dataset, our model achieved a precision of 76.59% and a mean Average Precision (mAP) of 64.15%, demonstrating competitive performance in detection accuracy, especially of small and low-contrast objects.
- 21 November 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15  (1) , 41347
Open AccessA multi-head YOLOv12 with self-supervised pretraining for urinary sediment particle detection
Automatic and reliable urine sediment analysis is essential for timely diagnosis and management of renal and urinary disorders. However, manual methods are time-consuming, subjective, and limited by operator abilities. In this study, we propose a novel deep learning method based on a multi-head YOLOv12 architecture combined with self-supervised pretraining and advanced inference through Slicing Aided Hyper Inference (SAHI) to effectively address these challenges. Unlike prior methods that employed a single detection head, our architecture features six specialized and independent detection heads: Cells, Casts, Crystals, Microorganisms/Yeast, Artifact, and Others, enabling simultaneous and fine-grained classification of the full spectrum of urine sediment particles, including all relevant subclasses. To facilitate robust training, we created a large-scale dataset (OpenUrine) encompassing 790 labeled images with over 31,285 bounding boxes across 39 categories, and 5640 unlabeled images for self-supervised learning. Evaluated on this complex 39-class dataset, our model achieved a precision of 76.59% and a mean Average Precision (mAP) of 64.15%, demonstrating competitive performance in detection accuracy, especially of small and low-contrast objects.
- Open Access
- 12 September 2025
- Published by eLife Sciences Publications, Ltd in eLife
- Vol. 13
The spatial frequency representation predicts category coding in the inferior temporal cortex
Understanding the neural representation of spatial frequency (SF) in the primate cortex is vital for unraveling visual processing in object recognition. While many studies focus on SF in the primary visual cortex, the characteristics and interaction of SF with category representation remain unclear. To explore SF representation in the inferior temporal (IT) cortex of macaques, we conducted extracellular recordings with complex stimuli systematically filtered by SF. Our findings reveal explicit SF coding at both single-neuron and population levels. Temporal dynamics analysis of SF representation reveals that low SF (LSF) is decoded faster than high SF (HSF), and the SF preference dynamically shifts from LSF to HSF over time. Additionally, the SF representation for each neuron forms a profile that predicts category selectivity at the population level. IT neurons cluster into four groups based on SF preference, each with distinct category coding behaviors. Notably, HSF-preferring neurons show the highest category decoding for faces. Despite the existing connection between SF and category coding, we have identified uncorrelated representations of SF and category. Unlike category coding, SF is more sparsely represented and depends more on individual neurons. These findings dissociate SF and category representations, underscoring SF’s pivotal role in object recognition.
- 12 September 2025
- Published by eLife Sciences Publications, Ltd in eLife
- Vol. 13
Open AccessThe spatial frequency representation predicts category coding in the inferior temporal cortex
Understanding the neural representation of spatial frequency (SF) in the primate cortex is vital for unraveling visual processing in object recognition. While many studies focus on SF in the primary visual cortex, the characteristics and interaction of SF with category representation remain unclear. To explore SF representation in the inferior temporal (IT) cortex of macaques, we conducted extracellular recordings with complex stimuli systematically filtered by SF. Our findings reveal explicit SF coding at both single-neuron and population levels. Temporal dynamics analysis of SF representation reveals that low SF (LSF) is decoded faster than high SF (HSF), and the SF preference dynamically shifts from LSF to HSF over time. Additionally, the SF representation for each neuron forms a profile that predicts category selectivity at the population level. IT neurons cluster into four groups based on SF preference, each with distinct category coding behaviors. Notably, HSF-preferring neurons show the highest category decoding for faces. Despite the existing connection between SF and category coding, we have identified uncorrelated representations of SF and category. Unlike category coding, SF is more sparsely represented and depends more on individual neurons. These findings dissociate SF and category representations, underscoring SF’s pivotal role in object recognition.
- Open Access
- 14 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Systems and Rehabilitation Engineering
- Vol. PP  (15344320) , 1
DiffuseGaitNet: Improving Parkinson’s Disease Gait Severity Assessment with a Diffusion Model Framework
Assessing the severity of gait impairment in Parkinson’s disease (PD) using the Movement Disorder Society’s Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) is typically performed by clinical experts, but this process is time-consuming, subjective, and costly. To address these challenges, we propose a Guided Diffusion Model with an encoder-only transformer that automatically predicts gait severity by learning the underlying distribution of PD gait and leveraging domain knowledge critical for clinical evaluations. Our diffusion model enables us to generate synthetic PD gait video frames conditioned on clinical features determined by experts to assess disease severity. These synthetic samples contain novel movement patterns not present in the observed data; systems trained on this information have better prediction performance. In addition, we propose a novel classification algorithm that can learn a predictive model, from both observed training data and synthetic samples, to accurately assess PD severity. We evaluate the effectiveness of the proposed method using two human motion datasets across two tasks: PD severity prediction and action classification. Our approach in predicting PD, and our action classification is sufficiently accurate that it can be applied to general applications with healthy subjects performing similar tasks. The full codebase is available on GitHub: https://github.com/arshakRz/DiffuseGaitNet.
- 14 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Systems and Rehabilitation Engineering
- Vol. PP  (15344320) , 1
Open AccessDiffuseGaitNet: Improving Parkinson’s Disease Gait Severity Assessment with a Diffusion Model Framework
Assessing the severity of gait impairment in Parkinson’s disease (PD) using the Movement Disorder Society’s Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) is typically performed by clinical experts, but this process is time-consuming, subjective, and costly. To address these challenges, we propose a Guided Diffusion Model with an encoder-only transformer that automatically predicts gait severity by learning the underlying distribution of PD gait and leveraging domain knowledge critical for clinical evaluations. Our diffusion model enables us to generate synthetic PD gait video frames conditioned on clinical features determined by experts to assess disease severity. These synthetic samples contain novel movement patterns not present in the observed data; systems trained on this information have better prediction performance. In addition, we propose a novel classification algorithm that can learn a predictive model, from both observed training data and synthetic samples, to accurately assess PD severity. We evaluate the effectiveness of the proposed method using two human motion datasets across two tasks: PD severity prediction and action classification. Our approach in predicting PD, and our action classification is sufficiently accurate that it can be applied to general applications with healthy subjects performing similar tasks. The full codebase is available on GitHub: https://github.com/arshakRz/DiffuseGaitNet.
- Open Access
- 24 May 2025
- Published by Springer Nature in Discover Applied Sciences
- Vol. 7  (6) , 1-24
Brand visibility in packaging: a deep learning approach for logo detection, saliency-map prediction, and logo placement analysis
- 24 May 2025
- Published by Springer Nature in Discover Applied Sciences
- Vol. 7  (6) , 1-24
Open AccessBrand visibility in packaging: a deep learning approach for logo detection, saliency-map prediction, and logo placement analysis
- Open Access
- 27 February 2025
- Published by eLife Sciences Publications, Ltd in eLife preprints
- Vol. 13
The Spatial Frequency Representation Predicts Category Coding in the Inferior Temporal Cortex
Understanding the neural representation of spatial frequency (SF) in the primate cortex is vital for unraveling visual processing mechanisms in object recognition. While numerous studies concentrate on the representation of SF in the primary visual cortex, the characteristics of SF representation and its interaction with category representation remain inadequately understood. To explore SF representation in the inferior temporal (IT) cortex of macaque monkeys, we conducted extracellular recordings with complex stimuli systematically filtered by SF. Our findings disclose an explicit SF coding at single-neuron and population levels in the IT cortex. Moreover, the coding of SF content exhibits a coarse-to-fine pattern, declining as the SF increases. Temporal dynamics analysis of SF representation reveals that low SF (LSF) is decoded faster than high SF (HSF), and the SF preference dynamically shifts from LSF to HSF over time. Additionally, the SF representation for each neuron forms a profile that predicts category selectivity at the population level. IT neurons can be clustered into four groups based on SF preference, each exhibiting different category coding behaviors. Particularly, HSF-preferred neurons demonstrate the highest category decoding performance for face stimuli. Despite the existing connection between SF and category coding, we have identified uncorrelated representations of SF and category. In contrast to the category coding, SF is more sparse and places greater reliance on the representations of individual neurons. Comparing SF representation in the IT cortex to deep neural networks, we observed no relationship between SF representation and category coding. However, SF coding, as a category-orthogonal property, is evident across various ventral stream models. These results dissociate the separate representations of SF and object category, underscoring the pivotal role of SF in object recognition.
- 27 February 2025
- Published by eLife Sciences Publications, Ltd in eLife preprints
- Vol. 13
Open AccessThe Spatial Frequency Representation Predicts Category Coding in the Inferior Temporal Cortex
Understanding the neural representation of spatial frequency (SF) in the primate cortex is vital for unraveling visual processing mechanisms in object recognition. While numerous studies concentrate on the representation of SF in the primary visual cortex, the characteristics of SF representation and its interaction with category representation remain inadequately understood. To explore SF representation in the inferior temporal (IT) cortex of macaque monkeys, we conducted extracellular recordings with complex stimuli systematically filtered by SF. Our findings disclose an explicit SF coding at single-neuron and population levels in the IT cortex. Moreover, the coding of SF content exhibits a coarse-to-fine pattern, declining as the SF increases. Temporal dynamics analysis of SF representation reveals that low SF (LSF) is decoded faster than high SF (HSF), and the SF preference dynamically shifts from LSF to HSF over time. Additionally, the SF representation for each neuron forms a profile that predicts category selectivity at the population level. IT neurons can be clustered into four groups based on SF preference, each exhibiting different category coding behaviors. Particularly, HSF-preferred neurons demonstrate the highest category decoding performance for face stimuli. Despite the existing connection between SF and category coding, we have identified uncorrelated representations of SF and category. In contrast to the category coding, SF is more sparse and places greater reliance on the representations of individual neurons. Comparing SF representation in the IT cortex to deep neural networks, we observed no relationship between SF representation and category coding. However, SF coding, as a category-orthogonal property, is evident across various ventral stream models. These results dissociate the separate representations of SF and object category, underscoring the pivotal role of SF in object recognition.
- 3 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Computer
- Vol. 57  (6) , 50-61
Gender Recognition Based on Hand Images Employing Local and Global Shape Information
Here, we propose a three-stream structure for hand gender recognition. The proposed method employs convolutional neural networks for feature extraction and classification parts to avoid any hand-designed feature extraction procedure. The performance of the proposed method is investigated by using a large hand dataset.
- 3 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Computer
- Vol. 57  (6) , 50-61
Gender Recognition Based on Hand Images Employing Local and Global Shape Information
Here, we propose a three-stream structure for hand gender recognition. The proposed method employs convolutional neural networks for feature extraction and classification parts to avoid any hand-designed feature extraction procedure. The performance of the proposed method is investigated by using a large hand dataset.
- Open Access
Brand Visibility in Packaging: A Deep Learning Approach for Logo Detection, Saliency-Map Prediction, and Logo Placement Analysis
In the highly competitive area of product marketing, the visibility of brand logos on packaging plays a crucial role in shaping consumer perception, directly influencing the success of the product. This paper introduces a comprehensive framework to measure the brand logo's attention on a packaging design. The proposed method consists of three steps. The first step leverages YOLOv8 for precise logo detection across prominent datasets, FoodLogoDet-1500 and LogoDet-3K. The second step involves modeling the user's visual attention with a novel saliency prediction model tailored for the packaging context. The proposed saliency model combines the visual elements with text maps employing a transformers-based architecture to predict user attention maps. In the third step, by integrating logo detection with a saliency map generation, the framework provides a comprehensive brand attention score. The effectiveness of the proposed method is assessed module by module, ensuring a thorough evaluation of each component. Comparing logo detection and saliency map prediction with state-of-the-art models shows the superiority of the proposed methods. To investigate the robustness of the proposed brand attention score, we collected a unique dataset to examine previous psychophysical hypotheses related to brand visibility. the results show that the brand attention score is in line with all previous studies. Also, we introduced seven new hypotheses to check the impact of position, orientation, presence of person, and other visual elements on brand attention. This research marks a significant stride in the intersection of cognitive psychology, computer vision, and marketing, paving the way for advanced, consumer-centric packaging designs.
Open AccessBrand Visibility in Packaging: A Deep Learning Approach for Logo Detection, Saliency-Map Prediction, and Logo Placement Analysis
In the highly competitive area of product marketing, the visibility of brand logos on packaging plays a crucial role in shaping consumer perception, directly influencing the success of the product. This paper introduces a comprehensive framework to measure the brand logo's attention on a packaging design. The proposed method consists of three steps. The first step leverages YOLOv8 for precise logo detection across prominent datasets, FoodLogoDet-1500 and LogoDet-3K. The second step involves modeling the user's visual attention with a novel saliency prediction model tailored for the packaging context. The proposed saliency model combines the visual elements with text maps employing a transformers-based architecture to predict user attention maps. In the third step, by integrating logo detection with a saliency map generation, the framework provides a comprehensive brand attention score. The effectiveness of the proposed method is assessed module by module, ensuring a thorough evaluation of each component. Comparing logo detection and saliency map prediction with state-of-the-art models shows the superiority of the proposed methods. To investigate the robustness of the proposed brand attention score, we collected a unique dataset to examine previous psychophysical hypotheses related to brand visibility. the results show that the brand attention score is in line with all previous studies. Also, we introduced seven new hypotheses to check the impact of position, orientation, presence of person, and other visual elements on brand attention. This research marks a significant stride in the intersection of cognitive psychology, computer vision, and marketing, paving the way for advanced, consumer-centric packaging designs.
- Open Access
- 1 January 2024
- Published by KMAN Publication Incorporation in Journal of Artificial Intelligence, Applications, and Innovations
- Vol. 1  (1) , 1-11
Revolutionizing Short Video Recommendations Using the Vision Mamba Framework
The rapid proliferation of short-form video content on platforms such as TikTok, Instagram, and YouTube Shorts has introduced significant challenges for recommendation systems, as traditional methods often struggle to keep up with the dynamic nature of user engagement and the large influx of data. In this paper, we present the Vision Mamba (Vim) framework, a cutting-edge approach in visual representation learning that employs bidirectional state space models to improve both the efficiency and accuracy of short video recommendations. The Vim framework excels by effectively capturing temporal dynamics, long-range dependencies, and the contextual relevance within video sequences, addressing computational limitations in a resource-efficient manner. Furthermore, it supports real-time personalization and scalable deployment across modern content platforms. Experimental evaluations conducted on the MicroLens dataset demonstrate that the Vision Mamba framework significantly outperforms existing traditional models, setting a new benchmark in video recommendation systems and offering enhanced user experiences with more contextually relevant and personalized content delivery.
- 1 January 2024
- Published by KMAN Publication Incorporation in Journal of Artificial Intelligence, Applications, and Innovations
- Vol. 1  (1) , 1-11
Open AccessRevolutionizing Short Video Recommendations Using the Vision Mamba Framework
The rapid proliferation of short-form video content on platforms such as TikTok, Instagram, and YouTube Shorts has introduced significant challenges for recommendation systems, as traditional methods often struggle to keep up with the dynamic nature of user engagement and the large influx of data. In this paper, we present the Vision Mamba (Vim) framework, a cutting-edge approach in visual representation learning that employs bidirectional state space models to improve both the efficiency and accuracy of short video recommendations. The Vim framework excels by effectively capturing temporal dynamics, long-range dependencies, and the contextual relevance within video sequences, addressing computational limitations in a resource-efficient manner. Furthermore, it supports real-time personalization and scalable deployment across modern content platforms. Experimental evaluations conducted on the MicroLens dataset demonstrate that the Vision Mamba framework significantly outperforms existing traditional models, setting a new benchmark in video recommendation systems and offering enhanced user experiences with more contextually relevant and personalized content delivery.