Shohreh Kasaei

Haniyeh Ehsani Oskouie, Mohammad-Shahram Moin, Shohreh Kasaei

Deep neural networks have made significant advancements in accurately estimating scene flow using point clouds, which is vital for many applications like video analysis, action recognition, and navigation. The robustness of these techniques, however, remains a concern, particularly in the face of adversarial attacks that have been proven to deceive state-of-the-art deep neural networks in many domains. Surprisingly, the robustness of scene flow networks against such attacks has not been thoroughly investigated. To address this problem, the proposed approach aims to bridge this gap by introducing adversarial white-box attacks specifically tailored for scene flow networks. Experimental results show that the generated adversarial examples obtain up to 33.7 relative degradation in average end-point error on the KITTI and FlyingThings3D datasets. The study also reveals the significant impact that attacks targeting point clouds in only one dimension or color channel have on average end-point error. Analyzing the success and failure of these attacks on the scene flow networks and their 2D optical flow network variants shows a higher vulnerability for the optical flow networks. Code is available at https://github.com/aheldis/Attack-on-Scene-Flow-using-Point-Clouds.git.

Amir M. Mansourian, Rozhan Ahmadi, Shohreh Kasaei

In recent years, deep neural networks have achieved remarkable accuracy in computer vision tasks. With inference time being a crucial factor, particularly in dense prediction tasks such as semantic segmentation, knowledge distillation has emerged as a successful technique for improving the accuracy of lightweight student networks. The existing methods often neglect the information in channels and among different classes. To overcome these limitations, this paper proposes a novel method called Inter-Class Similarity Distillation (ICSD) for the purpose of knowledge distillation. The proposed method transfers high-order relations from the teacher network to the student network by independently computing intra-class distributions for each class from network outputs. This is followed by calculating inter-class similarity matrices for distillation using KL divergence between distributions of each pair of classes. To further improve the effectiveness of the proposed method, an Adaptive Loss Weighting (ALW) training strategy is proposed. Unlike existing methods, the ALW strategy gradually reduces the influence of the teacher network towards the end of training process to account for errors in teacher's predictions. Extensive experiments conducted on two well-known datasets for semantic segmentation, Cityscapes and Pascal VOC 2012, validate the effectiveness of the proposed method in terms of mIoU and pixel accuracy. The proposed method outperforms most of existing knowledge distillation methods as demonstrated by both quantitative and qualitative evaluations. Code is available at: https://github.com/AmirMansurian/AICSD

Arash Chaichi Mellatshahi, Shohreh Kasaei

In recent years, limited research has discussed the loss function in the super-resolution process. The majority of those studies have only used perceptual similarity conventionally. This is while the development of appropriate loss can improve the quality of other methods as well. In this article, a new weighting method for pixel-wise loss is proposed. With the help of this method, it is possible to use trainable weights based on the general structure of the image and its perceptual features while maintaining the advantages of pixel-wise loss. Also, a criterion for comparing weights of loss is introduced so that the weights can be estimated directly by a convolutional neural network. In addition, in this article, the expectation-maximization method is used for the simultaneous estimation super-resolution network and weighting network. In addition, a new activation function, called "FixedSum", is introduced which can keep the sum of all components of vector constants while keeping the output components between zero and one. As experimental results shows, weighted loss by the proposed method leads to better results than the unweighted loss and weighted loss based on uncertainty in both signal-to-noise and perceptual similarity senses on the state-of-the-art networks. Code is available online.

Farnoosh Arefi, Amir M. Mansourian, Shohreh Kasaei

The performance of Video Instance Segmentation (VIS) methods has improved significantly with the advent of transformer networks. However, these networks often face challenges in training due to the high annotation cost. To address this, unsupervised and weakly-supervised methods have been developed to reduce the dependency on annotations. This work introduces a novel weakly-supervised method called Eigen-Cluster VIS that, without requiring any mask annotations, achieves competitive accuracy compared to other VIS approaches. This method is based on two key innovations: a Temporal Eigenvalue Loss (TEL) and a clip-level Quality Cluster Coefficient (QCC). The TEL ensures temporal coherence by leveraging the eigenvalues of the Laplacian matrix derived from graph adjacency matrices. By minimizing the mean absolute error between the eigenvalues of adjacent frames, this loss function promotes smooth transitions and stable segmentation boundaries over time, reducing temporal discontinuities and improving overall segmentation quality. The QCC employs the K-means method to ensure the quality of spatio-temporal clusters without relying on ground truth masks. Using the Davies-Bouldin score, the QCC provides an unsupervised measure of feature discrimination, allowing the model to self-evaluate and adapt to varying object distributions, enhancing robustness during the testing phase. These enhancements are computationally efficient and straightforward, offering significant performance gains without additional annotated data. The proposed Eigen-Cluster VIS method is evaluated on the YouTube-Video Instance Segmentation (YouTube-VIS) 2019/2021 and Occluded Video Instance Segmentation (OVIS) datasets, demonstrating that it effectively narrows the performance gap between the fully-supervised and weakly-supervised VIS approaches. The code is available on https://github.com/farnooshar/EigenClusterVIS

Hanieh Naderi, Chinthaka Dinesh, Ivan V. Bajic et al.

Adversarial attacks pose serious challenges for deep neural network (DNN)-based analysis of various input signals. In the case of three-dimensional point clouds, methods have been developed to identify points that play a key role in network decision, and these become crucial in generating existing adversarial attacks. For example, a saliency map approach is a popular method for identifying adversarial drop points, whose removal would significantly impact the network decision. This paper seeks to enhance the understanding of three-dimensional adversarial attacks by exploring which point cloud features are most important for predicting adversarial points. Specifically, Fourteen key point cloud features such as edge intensity and distance from the centroid are defined, and multiple linear regression is employed to assess their predictive power for adversarial points. Based on critical feature selection insights, a new attack method has been developed to evaluate whether the selected features can generate an attack successfully. Unlike traditional attack methods that rely on model-specific vulnerabilities, this approach focuses on the intrinsic characteristics of the point clouds themselves. It is demonstrated that these features can predict adversarial points across four different DNN architectures, Point Network (PointNet), PointNet++, Dynamic Graph Convolutional Neural Networks (DGCNN), and Point Convolutional Network (PointConv) outperforming random guessing and achieving results comparable to saliency map-based attacks. This study has important engineering applications, such as enhancing the security and robustness of three-dimensional point cloud-based systems in fields like robotics and autonomous driving.

Vladimir Somers, Victor Joos, Anthony Cioppa et al.

Tracking and identifying athletes on the pitch holds a central role in collecting essential insights from the game, such as estimating the total distance covered by players or understanding team tactics. This tracking and identification process is crucial for reconstructing the game state, defined by the athletes' positions and identities on a 2D top-view of the pitch, (i.e. a minimap). However, reconstructing the game state from videos captured by a single camera is challenging. It requires understanding the position of the athletes and the viewpoint of the camera to localize and identify players within the field. In this work, we formalize the task of Game State Reconstruction and introduce SoccerNet-GSR, a novel Game State Reconstruction dataset focusing on football videos. SoccerNet-GSR is composed of 200 video sequences of 30 seconds, annotated with 9.37 million line points for pitch localization and camera calibration, as well as over 2.36 million athlete positions on the pitch with their respective role, team, and jersey number. Furthermore, we introduce GS-HOTA, a novel metric to evaluate game state reconstruction methods. Finally, we propose and release an end-to-end baseline for game state reconstruction, bootstrapping the research on this task. Our experiments show that GSR is a challenging novel task, which opens the field for future research. Our dataset and codebase are publicly available at https://github.com/SoccerNet/sn-gamestate.

Amir M. Mansourian, Rozhan Ahmadi, Masoud Ghafouri et al.

Deep Neural Networks (DNNs) have achieved notable performance in the fields of computer vision and natural language processing with various applications in both academia and industry. However, with recent advancements in DNNs and transformer models with a tremendous number of parameters, deploying these large models on edge devices causes serious issues such as high runtime and memory consumption. This is especially concerning with the recent large-scale foundation models, Vision-Language Models (VLMs), and Large Language Models (LLMs). Knowledge Distillation (KD) is one of the prominent techniques proposed to address the aforementioned problems using a teacher-student architecture. More specifically, a lightweight student model is trained using additional knowledge from a cumbersome teacher model. In this work, a comprehensive survey of knowledge distillation methods is proposed. This includes reviewing KD from different aspects: distillation sources, distillation schemes, distillation algorithms, distillation by modalities, applications of distillation, and comparison among existing methods. In contrast to most existing surveys, which are either outdated or simply update former surveys, this work proposes a comprehensive survey with a new point of view and representation structure that categorizes and investigates the most recent methods in knowledge distillation. This survey considers various critically important subcategories, including KD for diffusion models, 3D inputs, foundational models, transformers, and LLMs. Furthermore, existing challenges in KD and possible future research directions are discussed. Github page of the project: https://github.com/IPL-Sharif/KD_Survey

Sina Elahimanesh, Mohammadali Mohammadkhani, Shohreh Kasaei

In contemporary society, social media is deeply integrated into daily life, yet emotional expression often differs between real and online contexts. We studied the Persian community on X to explore this gap, designing a human-centered pipeline to measure alignment between real-world and social media emotions. Recent tweets and images of participants were collected and analyzed using Transformers-based text and image sentiment modules. Friends of participants provided insights into their real-world emotions, which were compared with online expressions using a distance criterion. The study involved N=105 participants, 393 friends, over 8,300 tweets, and 2,000 media images. Results showed only 28% similarity between images and real-world emotions, while tweets aligned about 76% with participants' real-life feelings. Statistical analyses confirmed significant disparities in sentiment proportions across images, tweets, and friends' perceptions, highlighting differences in emotional expression between online and offline environments and demonstrating practical utility of the proposed pipeline for understanding digital self-presentation.

Navid Eslami, Farnoosh Arefi, Amir M. Mansourian et al.

Despite significant progress in deep learning-based optical flow methods, accurately estimating large displacements and repetitive patterns remains a challenge. The limitations of local features and similarity search patterns used in these algorithms contribute to this issue. Additionally, some existing methods suffer from slow runtime and excessive graphic memory consumption. To address these problems, this paper proposes a novel approach based on the RAFT framework. The proposed Attention-based Feature Localization (AFL) approach incorporates the attention mechanism to handle global feature extraction and address repetitive patterns. It introduces an operator for matching pixels with corresponding counterparts in the second frame and assigning accurate flow values. Furthermore, an Amorphous Lookup Operator (ALO) is proposed to enhance convergence speed and improve RAFTs ability to handle large displacements by reducing data redundancy in its search operator and expanding the search space for similarity extraction. The proposed method, Efficient RAFT (Ef-RAFT),achieves significant improvements of 10% on the Sintel dataset and 5% on the KITTI dataset over RAFT. Remarkably, these enhancements are attained with a modest 33% reduction in speed and a mere 13% increase in memory usage. The code is available at: https://github.com/n3slami/Ef-RAFT

Farnoosh Arefi, Amir M. Mansourian, Shohreh Kasaei

Deep spectral methods reframe the image decomposition process as a graph partitioning task by extracting features using self-supervised learning and utilizing the Laplacian of the affinity matrix to obtain eigensegments. However, instance segmentation has received less attention compared to other tasks within the context of deep spectral methods. This paper addresses the fact that not all channels of the feature map extracted from a self-supervised backbone contain sufficient information for instance segmentation purposes. In fact, Some channels are noisy and hinder the accuracy of the task. To overcome this issue, this paper proposes two channel reduction modules: Noise Channel Reduction (NCR) and Deviation-based Channel Reduction (DCR). The NCR retains channels with lower entropy, as they are less likely to be noisy, while DCR prunes channels with low standard deviation, as they lack sufficient information for effective instance segmentation. Furthermore, the paper demonstrates that the dot product, commonly used in deep spectral methods, is not suitable for instance segmentation due to its sensitivity to feature map values, potentially leading to incorrect instance segments. A new similarity metric called Bray-Curtis over Chebyshev (BoC) is proposed to address this issue. It takes into account the distribution of features in addition to their values, providing a more robust similarity measure for instance segmentation. Quantitative and qualitative results on the Youtube-VIS2019 dataset highlight the improvements achieved by the proposed channel reduction methods and the use of BoC instead of the conventional dot product for creating the affinity matrix. These improvements are observed in terms of mean Intersection over Union and extracted instance segments, demonstrating enhanced instance segmentation performance. The code is available on: https://github.com/farnooshar/SpecUnIIS

Rozhan Ahmadi, Shohreh Kasaei

In recent years, weakly supervised semantic segmentation using image-level labels as supervision has received significant attention in the field of computer vision. Most existing methods have addressed the challenges arising from the lack of spatial information in these labels by focusing on facilitating supervised learning through the generation of pseudo-labels from class activation maps (CAMs). Due to the localized pattern detection of CNNs, CAMs often emphasize only the most discriminative parts of an object, making it challenging to accurately distinguish foreground objects from each other and the background. Recent studies have shown that Vision Transformer (ViT) features, due to their global view, are more effective in capturing the scene layout than CNNs. However, the use of hierarchical ViTs has not been extensively explored in this field. This work explores the use of Swin Transformer by proposing "SWTformer" to enhance the accuracy of the initial seed CAMs by bringing local and global views together. SWTformer-V1 generates class probabilities and CAMs using only the patch tokens as features. SWTformer-V2 incorporates a multi-scale feature fusion mechanism to extract additional information and utilizes a background-aware mechanism to generate more accurate localization maps with improved cross-object discrimination. Based on experiments on the PascalVOC 2012 dataset, SWTformer-V1 achieves a 0.98% mAP higher localization accuracy, outperforming state-of-the-art models. It also yields comparable performance by 0.82% mIoU on average higher than other methods in generating initial localization maps, depending only on the classification network. SWTformer-V2 further improves the accuracy of the generated seed CAMs by 5.32% mIoU, further proving the effectiveness of the local-to-global view provided by the Swin transformer. Code available at: https://github.com/RozhanAhmadi/SWTformer

Haniyeh Ehsani Oskouie, Mohammad-Shahram Moin, Shohreh Kasaei

Deep neural networks have made significant advancements in accurately estimating scene flow using point clouds, which is vital for many applications like video analysis, action recognition, and navigation. The robustness of these techniques, however, remains a concern, particularly in the face of adversarial attacks that have been proven to deceive state-of-the-art deep neural networks in many domains. Surprisingly, the robustness of scene flow networks against such attacks has not been thoroughly investigated. To address this problem, the proposed approach aims to bridge this gap by introducing adversarial white-box attacks specifically tailored for scene flow networks. Experimental results show that the generated adversarial examples obtain up to 33.7 relative degradation in average end-point error on the KITTI and FlyingThings3D datasets. The study also reveals the significant impact that attacks targeting point clouds in only one dimension or color channel have on average end-point error. Analyzing the success and failure of these attacks on the scene flow networks and their 2D optical flow network variants shows a higher vulnerability for the optical flow networks. Code is available at https://github.com/aheldis/Attack-on-Scene-Flow-using-Point-Clouds.git.

Amir M. Mansourian, Amir Mohammad Babaei, Shohreh Kasaei

Multi-teacher knowledge distillation (KD), a more effective technique than traditional single-teacher methods, transfers knowledge from expert teachers to a compact student model using logit or feature matching. However, most existing approaches lack knowledge diversity, as they rely solely on unimodal visual information, overlooking the potential of cross-modal representations. In this work, we explore the use of CLIP's vision-language knowledge as a complementary source of supervision for KD, an area that remains largely underexplored. We propose a simple yet effective framework that fuses the logits and features of a conventional teacher with those from CLIP. By incorporating CLIP's multi-prompt textual guidance, the fused supervision captures both dataset-specific and semantically enriched visual cues. Beyond accuracy, analysis shows that the fused teacher yields more confident and reliable predictions, significantly increasing confident-correct cases while reducing confidently wrong ones. Moreover, fusion with CLIP refines the entire logit distribution, producing semantically meaningful probabilities for non-target classes, thereby improving inter-class consistency and distillation quality. Despite its simplicity, the proposed method, Enriching Knowledge Distillation (RichKD), consistently outperforms most existing baselines across multiple benchmarks and exhibits stronger robustness under distribution shifts and input corruptions.

AmirHossein Naghi Razlighi, Elaheh Badali Golezani, Shohreh Kasaei

3D Gaussian Splatting enables high-quality real-time rendering but often produces millions of splats, resulting in excessive storage and computational overhead. We propose a novel lossy compression method based on learnable confidence scores modeled as Beta distributions. Each splat's confidence is optimized through reconstruction-aware losses, enabling pruning of low-confidence splats while preserving visual fidelity. The proposed approach is architecture-agnostic and can be applied to any Gaussian Splatting variant. In addition, the average confidence values serve as a new metric to assess the quality of the scene. Extensive experiments demonstrate favorable trade-offs between compression and fidelity compared to prior work. Our code and data are publicly available at https://github.com/amirhossein-razlighi/Confident-Splatting

Amir M. Mansourian, Arya Jalali, Rozhan Ahmadi et al.

Deep learning models have achieved significant results across various computer vision tasks. However, due to the large number of parameters in these models, deploying them in real-time scenarios is a critical challenge, specifically in dense prediction tasks such as semantic segmentation. Knowledge distillation has emerged as a successful technique for addressing this problem by transferring knowledge from a cumbersome model (teacher) to a lighter model (student). In contrast to existing complex methodologies commonly employed for distilling knowledge from a teacher to a student, this paper showcases the efficacy of a simple yet powerful method for utilizing refined feature maps to transfer attention. The proposed method has proven to be effective in distilling rich information, outperforming existing methods in semantic segmentation as a dense prediction task. The proposed Attention-guided Feature Distillation (AttnFD) method, employs the Convolutional Block Attention Module (CBAM), which refines feature maps by taking into account both channel-specific and spatial information content. Simply using the Mean Squared Error (MSE) loss function between the refined feature maps of the teacher and the student, AttnFD demonstrates outstanding performance in semantic segmentation, achieving state-of-the-art results in terms of improving the mean Intersection over Union (mIoU) of the student network on the PascalVoc 2012, Cityscapes, COCO, and CamVid datasets.

Sina Elahimanesh, Mohammadali Mohammadkhani, Shohreh Kasaei

In contemporary society, widespread social media usage is evident in people's daily lives. Nevertheless, disparities in emotional expressions between the real world and online platforms can manifest. We comprehensively analyzed Persian community on X to explore this phenomenon. An innovative pipeline was designed to measure the similarity between emotions in the real world compared to social media. Accordingly, recent tweets and images of participants were gathered and analyzed using Transformers-based text and image sentiment analysis modules. Each participant's friends also provided insights into the their real-world emotions. A distance criterion was used to compare real-world feelings with virtual experiences. Our study encompassed N=105 participants, 393 friends who contributed their perspectives, over 8,300 collected tweets, and 2,000 media images. Results indicated a 28.67% similarity between images and real-world emotions, while tweets exhibited a 75.88% alignment with real-world feelings. Additionally, the statistical significance confirmed that the observed disparities in sentiment proportions.

Amir M. Mansourian, Vladimir Somers, Christophe De Vleeschouwer et al.

Effective tracking and re-identification of players is essential for analyzing soccer videos. But, it is a challenging task due to the non-linear motion of players, the similarity in appearance of players from the same team, and frequent occlusions. Therefore, the ability to extract meaningful embeddings to represent players is crucial in developing an effective tracking and re-identification system. In this paper, a multi-purpose part-based person representation method, called PRTreID, is proposed that performs three tasks of role classification, team affiliation, and re-identification, simultaneously. In contrast to available literature, a single network is trained with multi-task supervision to solve all three tasks, jointly. The proposed joint method is computationally efficient due to the shared backbone. Also, the multi-task learning leads to richer and more discriminative representations, as demonstrated by both quantitative and qualitative results. To demonstrate the effectiveness of PRTreID, it is integrated with a state-of-the-art tracking method, using a part-based post-processing module to handle long-term tracking. The proposed tracking method outperforms all existing tracking methods on the challenging SoccerNet tracking dataset.

Hanieh Naderi, Kimia Noorbakhsh, Arian Etemadi et al.

Although 3D point cloud classification has recently been widely deployed in different application scenarios, it is still very vulnerable to adversarial attacks. This increases the importance of robust training of 3D models in the face of adversarial attacks. Based on our analysis on the performance of existing adversarial attacks, more adversarial perturbations are found in the mid and high-frequency components of input data. Therefore, by suppressing the high-frequency content in the training phase, the models robustness against adversarial examples is improved. Experiments showed that the proposed defense method decreases the success rate of six attacks on PointNet, PointNet++ ,, and DGCNN models. In particular, improvements are achieved with an average increase of classification accuracy by 3.8 % on drop100 attack and 4.26 % on drop200 attack compared to the state-of-the-art methods. The method also improves models accuracy on the original dataset compared to other available methods.

Hassan Hafez-Kolahi, Behrad Moniri, Shohreh Kasaei

Two main concepts studied in machine learning theory are generalization gap (difference between train and test error) and excess risk (difference between test error and the minimum possible error). While information-theoretic tools have been used extensively to study the generalization gap of learning algorithms, the information-theoretic nature of excess risk has not yet been fully investigated. In this paper, some steps are taken toward this goal. We consider the frequentist problem of minimax excess risk as a zero-sum game between the algorithm designer and the world. Then, we argue that it is desirable to modify this game in a way that the order of play can be swapped. We then prove that, under some regularity conditions, if the world and designer can play randomly the duality gap is zero and the order of play can be changed. In this case, a Bayesian problem surfaces in the dual representation. This makes it possible to utilize recent information-theoretic results on minimum excess risk in Bayesian learning to provide bounds on the minimax excess risk. We demonstrate the applicability of the results by providing information theoretic insight on two important classes of problems: classification when the hypothesis space has finite VC-dimension, and regularized least squares.

Atrin Arya, Hanieh Naderi, Shohreh Kasaei

Recent research has revealed that the security of deep neural networks that directly process 3D point clouds to classify objects can be threatened by adversarial samples. Although existing adversarial attack methods achieve high success rates, they do not restrict the point modifications enough to preserve the point cloud appearance. To overcome this shortcoming, two constraints are proposed. These include applying hard boundary constraints on the number of modified points and on the point perturbation norms. Due to the restrictive nature of the problem, the search space contains many local maxima. The proposed method addresses this issue by using a high step-size at the beginning of the algorithm to search the main surface of the point cloud fast and effectively. Then, in order to converge to the desired output, the step-size is gradually decreased. To evaluate the performance of the proposed method, it is run on the ModelNet40 and ScanObjectNN datasets by employing the state-of-the-art point cloud classification models; including PointNet, PointNet++, and DGCNN. The obtained results show that it can perform successful attacks and achieve state-of-the-art results by only a limited number of point modifications while preserving the appearance of the point cloud. Moreover, due to the effective search algorithm, it can perform successful attacks in just a few steps. Additionally, the proposed step-size scheduling algorithm shows an improvement of up to $14.5\%$ when adopted by other methods as well. The proposed method also performs effectively against popular defense methods.

Seyed Mojtaba Marvasti-Zadeh, Javad Khaghani, Li Cheng et al.

A strong visual object tracker nowadays relies on its well-crafted modules, which typically consist of manually-designed network architectures to deliver high-quality tracking results. Not surprisingly, the manual design process becomes a particularly challenging barrier, as it demands sufficient prior experience, enormous effort, intuition, and perhaps some good luck. Meanwhile, neural architecture search has gaining grounds in practical applications as a promising method in tackling the issue of automated search of feasible network structures. In this work, we propose a novel cell-level differentiable architecture search mechanism with early stopping to automate the network design of the tracking module, aiming to adapt backbone features to the objective of Siamese tracking networks during offline training. Besides, the proposed early stopping strategy avoids over-fitting and performance collapse problems leading to generalization improvement. The proposed approach is simple, efficient, and with no need to stack a series of modules to construct a network. Our approach is easy to be incorporated into existing trackers, which is empirically validated using different differentiable architecture search-based methods and tracking objectives. Extensive experimental evaluations demonstrate the superior performance of our approach over five commonly-used benchmarks.

Hassan Hafez-Kolahi, Behrad Moniri, Shohreh Kasaei et al.

In parametric Bayesian learning, a prior is assumed on the parameter $W$ which determines the distribution of samples. In this setting, Minimum Excess Risk (MER) is defined as the difference between the minimum expected loss achievable when learning from data and the minimum expected loss that could be achieved if $W$ was observed. In this paper, we build upon and extend the recent results of (Xu & Raginsky, 2020) to analyze the MER in Bayesian learning and derive information-theoretic bounds on it. We formulate the problem as a (constrained) rate-distortion optimization and show how the solution can be bounded above and below by two other rate-distortion functions that are easier to study. The lower bound represents the minimum possible excess risk achievable by any process using $R$ bits of information from the parameter $W$. For the upper bound, the optimization is further constrained to use $R$ bits from the training set, a setting which relates MER to information-theoretic bounds on the generalization gap in frequentist learning. We derive information-theoretic bounds on the difference between these upper and lower bounds and show that they can provide order-wise tight rates for MER under certain conditions. This analysis gives more insight into the information-theoretic nature of Bayesian learning as well as providing novel bounds.

Hanieh Naderi, Leili Goli, Shohreh Kasaei

Deep neural networks have been shown to be vulnerable to adversarial examples deliberately constructed to misclassify victim models. As most adversarial examples have restricted their perturbations to $L_{p}$-norm, existing defense methods have focused on these types of perturbations and less attention has been paid to unrestricted adversarial examples; which can create more realistic attacks, able to deceive models without affecting human predictions. To address this problem, the proposed adversarial attack generates an unrestricted adversarial example with a limited number of parameters. The attack selects three points on the input image and based on their locations transforms the image into an adversarial example. By limiting the range of movement and location of these three points and using a discriminatory network, the proposed unrestricted adversarial example preserves the image appearance. Experimental results show that the proposed adversarial examples obtain an average success rate of 93.5% in terms of human evaluation on the MNIST and SVHN datasets. It also reduces the model accuracy by an average of 73% on six datasets MNIST, FMNIST, SVHN, CIFAR10, CIFAR100, and ImageNet. It should be noted that, in the case of attacks, lower accuracy in the victim model denotes a more successful attack. The adversarial train of the attack also improves model robustness against a randomly transformed image.

Mahdi Ghorbani, Fahimeh Fooladgar, Shohreh Kasaei

Deep neural network architectures have attained remarkable improvements in scene understanding tasks. Utilizing an efficient model is one of the most important constraints for limited-resource devices. Recently, several compression methods have been proposed to diminish the heavy computational burden and memory consumption. Among them, the pruning and quantizing methods exhibit a critical drop in performances by compressing the model parameters. While the knowledge distillation methods improve the performance of compact models by focusing on training lightweight networks with the supervision of cumbersome networks. In the proposed method, the knowledge distillation has been performed within the network by constructing multiple branches over the primary stream of the model, known as the self-distillation method. Therefore, the ensemble of sub-neural network models has been proposed to transfer the knowledge among themselves with the knowledge distillation policies as well as an adversarial learning strategy. Hence, The proposed ensemble of sub-models is trained against a discriminator model adversarially. Besides, their knowledge is transferred within the ensemble by four different loss functions. The proposed method has been devoted to both lightweight image classification and encoder-decoder architectures to boost the performance of small and compact models without incurring extra computational overhead at the inference process. Extensive experimental results on the main challenging datasets show that the proposed network outperforms the primary model in terms of accuracy at the same number of parameters and computational cost. The obtained results show that the proposed model has achieved significant improvement over earlier ideas of self-distillation methods. The effectiveness of the proposed models has also been illustrated in the encoder-decoder model.

Seyed Mojtaba Marvasti-Zadeh, Hossein Ghanei-Yakhdan, Shohreh Kasaei

Due to the automatic feature extraction procedure via multi-layer nonlinear transformations, the deep learning-based visual trackers have recently achieved great success in challenging scenarios for visual tracking purposes. Although many of those trackers utilize the feature maps from pre-trained convolutional neural networks (CNNs), the effects of selecting different models and exploiting various combinations of their feature maps are still not compared completely. To the best of our knowledge, all those methods use a fixed number of convolutional feature maps without considering the scene attributes (e.g., occlusion, deformation, and fast motion) that might occur during tracking. As a pre-requisition, this paper proposes adaptive discriminative correlation filters (DCF) based on the methods that can exploit CNN models with different topologies. First, the paper provides a comprehensive analysis of four commonly used CNN models to determine the best feature maps of each model. Second, with the aid of analysis results as attribute dictionaries, adaptive exploitation of deep features is proposed to improve the accuracy and robustness of visual trackers regarding video characteristics. Third, the generalization of the proposed method is validated on various tracking datasets as well as CNN models with similar architectures. Finally, extensive experimental results demonstrate the effectiveness of the proposed adaptive method compared with state-of-the-art visual tracking methods.

Seyed Mojtaba Marvasti-Zadeh, Javad Khaghani, Hossein Ghanei-Yakhdan et al.

We consider the problem of tracking an unknown small target from aerial videos of medium to high altitudes. This is a challenging problem, which is even more pronounced in unavoidable scenarios of drastic camera motion and high density. To address this problem, we introduce a context-aware IoU-guided tracker (COMET) that exploits a multitask two-stream network and an offline reference proposal generation strategy. The proposed network fully exploits target-related information by multi-scale feature learning and attention modules. The proposed strategy introduces an efficient sampling strategy to generalize the network on the target and its parts without imposing extra computational complexity during online tracking. These strategies contribute considerably in handling significant occlusions and viewpoint changes. Empirically, COMET outperforms the state-of-the-arts in a range of aerial view datasets that focusing on tracking small objects. Specifically, COMET outperforms the celebrated ATOM tracker by an average margin of 6.2% (and 7%) in precision (and success) score on challenging benchmarks of UAVDT, VisDrone-2019, and Small-90.

Seyed Mojtaba Marvasti-Zadeh, Hossein Ghanei-Yakhdan, Shohreh Kasaei

In recent years, visual tracking methods that are based on discriminative correlation filters (DCF) have been very promising. However, most of these methods suffer from a lack of robust scale estimation skills. Although a wide range of recent DCF-based methods exploit the features that are extracted from deep convolutional neural networks (CNNs) in their translation model, the scale of the visual target is still estimated by hand-crafted features. Whereas the exploitation of CNNs imposes a high computational burden, this paper exploits pre-trained lightweight CNNs models to propose two efficient scale estimation methods, which not only improve the visual tracking performance but also provide acceptable tracking speeds. The proposed methods are formulated based on either holistic or region representation of convolutional feature maps to efficiently integrate into DCF formulations to learn a robust scale model in the frequency domain. Moreover, against the conventional scale estimation methods with iterative feature extraction of different target regions, the proposed methods exploit proposed one-pass feature extraction processes that significantly improve the computational efficiency. Comprehensive experimental results on the OTB-50, OTB-100, TC-128 and VOT-2018 visual tracking datasets demonstrate that the proposed visual tracking methods outperform the state-of-the-art methods, effectively.

Seyed Mojtaba Marvasti-Zadeh, Hossein Ghanei-Yakhdan, Shohreh Kasaei

In recent years, the background-aware correlation filters have achie-ved a lot of research interest in the visual target tracking. However, these methods cannot suitably model the target appearance due to the exploitation of hand-crafted features. On the other hand, the recent deep learning-based visual tracking methods have provided a competitive performance along with extensive computations. In this paper, an adaptive background-aware correlation filter-based tracker is proposed that effectively models the target appearance by using either the histogram of oriented gradients (HOG) or convolutional neural network (CNN) feature maps. The proposed method exploits the fast 2D non-maximum suppression (NMS) algorithm and the semantic information comparison to detect challenging situations. When the HOG-based response map is not reliable, or the context region has a low semantic similarity with prior regions, the proposed method constructs the CNN context model to improve the target region estimation. Furthermore, the rejection option allows the proposed method to update the CNN context model only on valid regions. Comprehensive experimental results demonstrate that the proposed adaptive method clearly outperforms the accuracy and robustness of visual target tracking compared to the state-of-the-art methods on the OTB-50, OTB-100, TC-128, UAV-123, and VOT-2015 datasets.

Seyed Mojtaba Marvasti-Zadeh, Hossein Ghanei-Yakhdan, Shohreh Kasaei et al.

Visual object tracking remains an active research field in computer vision due to persisting challenges with various problem-specific factors in real-world scenes. Many existing tracking methods based on discriminative correlation filters (DCFs) employ feature extraction networks (FENs) to model the target appearance during the learning process. However, using deep feature maps extracted from FENs based on different residual neural networks (ResNets) has not previously been investigated. This paper aims to evaluate the performance of twelve state-of-the-art ResNet-based FENs in a DCF-based framework to determine the best for visual tracking purposes. First, it ranks their best feature maps and explores the generalized adoption of the best ResNet-based FEN into another DCF-based method. Then, the proposed method extracts deep semantic information from a fully convolutional FEN and fuses it with the best ResNet-based feature maps to strengthen the target representation in the learning process of continuous convolution filters. Finally, it introduces a new and efficient semantic weighting method (using semantic segmentation feature maps on each video frame) to reduce the drift problem. Extensive experimental results on the well-known OTB-2013, OTB-2015, TC-128 and VOT-2018 visual tracking datasets demonstrate that the proposed method effectively outperforms state-of-the-art methods in terms of precision and robustness of visual tracking.

Fahimeh Fooladgar, Shohreh Kasaei

Extremely efficient convolutional neural network architectures are one of the most important requirements for limited-resource devices (such as embedded and mobile devices). The computing power and memory size are two important constraints of these devices. Recently, some architectures have been proposed to overcome these limitations by considering specific hardware-software equipment. In this paper, the lightweight residual densely connected blocks are proposed to guaranty the deep supervision, efficient gradient flow, and feature reuse abilities of convolutional neural network. The proposed method decreases the cost of training and inference processes without using any special hardware-software equipment by just reducing the number of parameters and computational operations while achieving a feasible accuracy. Extensive experimental results demonstrate that the proposed architecture is more efficient than the AlexNet and VGGNet in terms of model size, required parameters, and even accuracy. The proposed model has been evaluated on the ImageNet, MNIST, Fashion MNIST, SVHN, CIFAR-10, and CIFAR-100. It achieves state-of-the-art results on Fashion MNIST dataset and reasonable results on the others. The obtained results show the superiority of the proposed method to efficient models such as the SqueezNet. It is also comparable with state-of-the-art efficient models such as CondenseNet and ShuffleNet.

Mobina Mahdavi, Fahimeh Fooladgar, Shohreh Kasaei

With the rapid progress of deep convolutional neural networks, in almost all robotic applications, the availability of 3D point clouds improves the accuracy of 3D semantic segmentation methods. Rendering of these irregular, unstructured, and unordered 3D points to 2D images from multiple viewpoints imposes some issues such as loss of information due to 3D to 2D projection, discretizing artifacts, and high computational costs. To efficiently deal with a large number of points and incorporate more context of each point, a pointwise attention-based atrous convolutional neural network architecture is proposed. It focuses on salient 3D feature points among all feature maps while considering outstanding contextual information via spatial channel-wise attention modules. The proposed model has been evaluated on the two most important 3D point cloud datasets for the 3D semantic segmentation task. It achieves a reasonable performance compared to state-of-the-art models in terms of accuracy, with a much smaller number of parameters.

Fahimeh Fooladgar, Shohreh Kasaei

The 3D scene understanding is mainly considered as a crucial requirement in computer vision and robotics applications. One of the high-level tasks in 3D scene understanding is semantic segmentation of RGB-Depth images. With the availability of RGB-D cameras, it is desired to improve the accuracy of the scene understanding process by exploiting the depth features along with the appearance features. As depth images are independent of illumination, they can improve the quality of semantic labeling alongside RGB images. Consideration of both common and specific features of these two modalities improves the performance of semantic segmentation. One of the main problems in RGB-Depth semantic segmentation is how to fuse or combine these two modalities to achieve more advantages of each modality while being computationally efficient. Recently, the methods that encounter deep convolutional neural networks have reached the state-of-the-art results by early, late, and middle fusion strategies. In this paper, an efficient encoder-decoder model with the attention-based fusion block is proposed to integrate mutual influences between feature maps of these two modalities. This block explicitly extracts the interdependences among concatenated feature maps of these modalities to exploit more powerful feature maps from RGB-Depth images. The extensive experimental results on three main challenging datasets of NYU-V2, SUN RGB-D, and Stanford 2D-3D-Semantic show that the proposed network outperforms the state-of-the-art models with respect to computational cost as well as model size. Experimental results also illustrate the effectiveness of the proposed lightweight attention-based fusion model in terms of accuracy.

Seyed Mojtaba Marvasti-Zadeh, Li Cheng, Hossein Ghanei-Yakhdan et al.

Visual target tracking is one of the most sought-after yet challenging research topics in computer vision. Given the ill-posed nature of the problem and its popularity in a broad range of real-world scenarios, a number of large-scale benchmark datasets have been established, on which considerable methods have been developed and demonstrated with significant progress in recent years -- predominantly by recent deep learning (DL)-based methods. This survey aims to systematically investigate the current DL-based visual tracking methods, benchmark datasets, and evaluation metrics. It also extensively evaluates and analyzes the leading visual tracking methods. First, the fundamental characteristics, primary motivations, and contributions of DL-based methods are summarized from nine key aspects of: network architecture, network exploitation, network training for visual tracking, network objective, network output, exploitation of correlation filter advantages, aerial-view tracking, long-term tracking, and online tracking. Second, popular visual tracking benchmarks and their respective properties are compared, and their evaluation metrics are summarized. Third, the state-of-the-art DL-based methods are comprehensively examined on a set of well-established benchmarks of OTB2013, OTB2015, VOT2018, LaSOT, UAV123, UAVDT, and VisDrone2019. Finally, by conducting critical analyses of these state-of-the-art trackers quantitatively and qualitatively, their pros and cons under various common scenarios are investigated. It may serve as a gentle use guide for practitioners to weigh when and under what conditions to choose which method(s). It also facilitates a discussion on ongoing issues and sheds light on promising research directions.

Hassan Hafez-Kolahi, Shohreh Kasaei, Mahdiyeh Soleymani-Baghshah

One of the most studied problems in machine learning is finding reasonable constraints that guarantee the generalization of a learning algorithm. These constraints are usually expressed as some simplicity assumptions on the target. For instance, in the Vapnik-Chervonenkis (VC) theory the space of possible hypotheses is considered to have a limited VC dimension. In this paper, the constraint on the entropy $H(X)$ of the input variable $X$ is studied as a simplicity assumption. It is proven that the sample complexity to achieve an $ε$-$δ$ Probably Approximately Correct (PAC) hypothesis is bounded by $\frac{2^{ \left.6H(X)\middle/ε\right.}+\log{\frac{1}δ}}{ε^2}$ which is sharp up to the $\frac{1}{ε^2}$ factor. Morever, it is shown that if a feature learning process is employed to learn the compressed representation from the dataset, this bound no longer exists. These findings have important implications on the Information Bottleneck (IB) theory which had been utilized to explain the generalization power of Deep Neural Networks (DNNs), but its applicability for this purpose is currently under debate by researchers. In particular, this is a rigorous proof for the previous heuristic that compressed representations are exponentially easier to be learned. However, our analysis pinpoints two factors preventing the IB, in its current form, to be applicable in studying neural networks. Firstly, the exponential dependence of sample complexity on $\frac{1}ε$, which can lead to a dramatic effect on the bounds in practical applications when $ε$ is small. Secondly, our analysis reveals that arguments based on input compression are inherently insufficient to explain generalization of methods like DNNs in which the features are also learned using available data.

Hassan Hafez-Kolahi, Shohreh Kasaei

Information Theory (IT) has been used in Machine Learning (ML) from early days of this field. In the last decade, advances in Deep Neural Networks (DNNs) have led to surprising improvements in many applications of ML. The result has been a paradigm shift in the community toward revisiting previous ideas and applications in this new framework. Ideas from IT are no exception. One of the ideas which is being revisited by many researchers in this new era, is Information Bottleneck (IB); a formulation of information extraction based on IT. The IB is promising in both analyzing and improving DNNs. The goal of this survey is to review the IB concept and demonstrate its applications in deep learning. The information theoretic nature of IB, makes it also a good candidate in showing the more general concept of how IT can be used in ML. Two important concepts are highlighted in this narrative on the subject, i) the concise and universal view that IT provides on seemingly unrelated methods of ML, demonstrated by explaining how IB relates to minimal sufficient statistics, stochastic gradient descent, and variational auto-encoders, and ii) the common technical mistakes and problems caused by applying ideas from IT, which is discussed by a careful study of some recent methods suffering from them.

Seyed Mojtaba Marvasti-Zadeh, Hossein Ghanei-Yakhdan, Shohreh Kasaei

With the fast growth of communication networks, the video data transmission from these networks is extremely vulnerable. Error concealment is a technique to estimate the damaged data by employing the correctly received data at the decoder. In this paper, an efficient boundary matching algorithm for estimating damaged motion vectors (MVs) is proposed. The proposed algorithm performs error concealment for each damaged macro block (MB) according to the list of identified priority of each frame. It then uses a classic boundary matching criterion or the proposed boundary matching criterion adaptively to identify matching distortion in each boundary of candidate MB. Finally, the candidate MV with minimum distortion is selected as an MV of damaged MB and the list of priorities is updated. Experimental results show that the proposed algorithm improves both objective and subjective qualities of reconstructed frames without any significant increase in computational cost. The PSNR for test sequences in some frames is increased about 4.7, 4.5, and 4.4 dB compared to the classic boundary matching, directional boundary matching, and directional temporal boundary matching algorithm, respectively.

Seyed Mojtaba Marvasti-Zadeh, Hossein Ghanei-Yakhdan, Shohreh Kasaei

Sending compressed video data in error-prone environments (like the Internet and wireless networks) might cause data degradation. Error concealment techniques try to conceal the received data in the decoder side. In this paper, an adaptive boundary matching algorithm is presented for recovering the damaged motion vectors (MVs). This algorithm uses an outer boundary matching or directional temporal boundary matching method to compare every boundary of candidate macroblocks (MBs), adaptively. It gives a specific weight according to the accuracy of each boundary of the damaged MB. Moreover, if each of the adjacent MBs is already concealed, different weights are given to the boundaries. Finally, the MV with minimum adaptive boundary distortion is selected as the MV of the damaged MB. Experimental results show that the proposed algorithm can improve both objective and subjective quality of reconstructed frames without any considerable computational complexity. The average PSNR in some frames of test sequences increases about 5.20, 5.78, 5.88, 4.37, 4.41, and 3.50 dB compared to average MV, classic boundary matching, directional boundary matching, directional temporal boundary matching, outer boundary matching, and dynamical temporal error concealment algorithm, respectively.

Behnam Babagholami-Mohamadabadi, Amin Jourabloo, Ali Zarghami et al.

A Bayesian framework for 3D human pose estimation from monocular images based on sparse representation (SR) is introduced. Our probabilistic approach aims at simultaneously learning two overcomplete dictionaries (one for the visual input space and the other for the pose space) with a shared sparse representation. Existing SR-based pose estimation approaches only offer a point estimation of the dictionary and the sparse codes. Therefore, they might be unreliable when the number of training examples is small. Our Bayesian framework estimates a posterior distribution for the sparse codes and the dictionaries from labeled training data. Hence, it is robust to overfitting on small-size training data. Experimental results on various human activities show that the proposed method is superior to the state of-the-art pose estimation algorithms.