Reza Hoseinnezhad

Bailu Wang, Wei Yi, Reza Hoseinnezhad et al.

In this paper, we propose a distributed multi-object tracking algorithm through the use of multi-Bernoulli (MB) filter based on generalized Covariance Intersection (G-CI). Our analyses show that the G-CI fusion with two MB posterior distributions does not admit an accurate closed-form expression. To solve this challenging problem, we firstly approximate the fused posterior as the unlabeled version of $δ$-generalized labeled multi-Bernoulli ($δ$-GLMB) distribution, referred to as generalized multi-Bernoulli (GMB) distribution. Then, to allow the subsequent fusion with another multi-Bernoulli posterior distribution, e.g., fusion with a third sensor node in the sensor network, or fusion in the feedback working mode, we further approximate the fused GMB posterior distribution as an MB distribution which matches its first-order statistical moment. The proposed fusion algorithm is implemented using sequential Monte Carlo technique and its performance is highlighted by numerical results.

Suqi Li, Wei Yi, Reza Hoseinnezhad et al.

This paper considers the problem of the distributed fusion of multi-object posteriors in the labeled random finite set filtering framework, using Generalized Covariance Intersection (GCI) method. Our analysis shows that GCI fusion with labeled multi-object densities strongly relies on label consistencies between local multi-object posteriors at different sensor nodes, and hence suffers from a severe performance degradation when perfect label consistencies are violated. Moreover, we mathematically analyze this phenomenon from the perspective of Principle of Minimum Discrimination Information and the so called yes-object probability. Inspired by the analysis, we propose a novel and general solution for the distributed fusion with labeled multi-object densities that is robust to label inconsistencies between sensors. Specifically, the labeled multi-object posteriors are firstly marginalized to their unlabeled posteriors which are then fused using GCI method. We also introduce a principled method to construct the labeled fused density and produce tracks formally. Based on the developed theoretical framework, we present tractable algorithms for the family of generalized labeled multi-Bernoulli (GLMB) filters including $δ$-GLMB, marginalized $δ$-GLMB and labeled multi-Bernoulli filters. The robustness and efficiency of the proposed distributed fusion algorithm are demonstrated in challenging tracking scenarios via numerical experiments.

Suqi Li, Wei Yi, Reza Hoseinnezhad et al.

This paper presents an exact Bayesian filtering solution for the multi-object tracking problem with the generic observation model. The proposed solution is designed in the labeled random finite set framework, using the product styled representation of labeled multi-object densities, with the standard multi-object transition kernel and no particular simplifying assumptions on the multi-object likelihood. Computationally tractable solutions are also devised by applying a principled approximation involving the replacement of the full multi-object density with a labeled multi-Bernoulli density that minimizes the Kullback-Leibler divergence and preserves the first-order moment. To achieve the fast performance, a dynamic grouping procedure based implementation is presented with a step-by-step algorithm. The performance of the proposed filter and its tractable implementations are verified and compared with the state-of-the-art in numerical experiments.

Meng Jiang, Wei Yi, Reza Hoseinnezhad et al.

The paper addresses distributed multi-target tracking in the framework of generalized Covariance Intersection (GCI) over multistatic radar system. The proposed method is based on the unlabeled version of generalized labeled multi-Bernoulli (GLMB) family by discarding the labels, referred as generalized multi-Bernoulli (GMB) family. However, it doesn't permit closed form solution for GCI fusion with GMB family. To solve this challenging problem, firstly, we propose an efficient approximation to the GMB family which preserves both the probability hypothesis density (PHD) and cardinality distribution, named as second-order approximation of GMB (SO-GMB) density. Then, we derive explicit expression for the GCI fusion with SO-GMB density. Finally, we compare the first-order approximation of GMB (FO-GMB) density with SO-GMB density in two scenarios and make a concrete analysis of the advantages of the second-order approximation. Simulation results are presented to verify the proposed approach.

Xiaoying Wang, Reza Hoseinnezhad, Amirali K. Gostar et al.

Sensor management in multi-object stochastic systems is a theoretically and computationally challenging problem. This paper presents a novel approach to the multi-target multi-sensor control problem within the partially observed Markov decision process (POMDP) framework. We model the multi-object state as a labeled multi-Bernoulli random finite set (RFS), and use the labeled multi-Bernoulli filter in conjunction with minimizing a task-driven control objective function: posterior expected error of cardinality and state (PEECS). A major contribution is a guided search for multi-dimensional optimization in the multi-sensor control command space, using coordinate descent method. In conjunction with the Generalized Covariance Intersection method for multi-sensor fusion, a fast multi-sensor algorithm is achieved. Numerical studies are presented in several scenarios where numerous controllable (mobile) sensors track multiple moving targets with different levels of observability. The results show that our method works significantly faster than the approach taken by a state of art method, with similar tracking errors.

Amirali K. Gostar, Reza Hoseinnezhad, Alireza Bab-Hadiashar

The recently developed labeled multi-Bernoulli (LMB) filter uses better approximations in its update step, compared to the unlabeled multi-Bernoulli filters, and more importantly, it provides us with not only the estimates for the number of targets and their states, but also with labels for existing tracks. This paper presents a novel sensor-control method to be used for optimal multi-target tracking within the LMB filter. The proposed method uses a task-driven cost function in which both the state estimation errors and cardinality estimation errors are taken into consideration. Simulation results demonstrate that the proposed method can successfully guide a mobile sensor in a challenging multi-target tracking scenario.

Nida Ishtiaq, Amirali Khodadadian Gostar, Alireza Bab-Hadiashar et al.

Tracking multiple objects through time is an important part of an intelligent transportation system. Random finite set (RFS)-based filters are one of the emerging techniques for tracking multiple objects. In multi-object tracking (MOT), a common assumption is that each object is moving independent of its surroundings. But in many real-world applications, target objects interact with one another and the environment. Such interactions, when considered for tracking, are usually modeled by an interactive motion model which is application specific. In this paper, we present a novel approach to incorporate target interactions within the prediction step of an RFS-based multi-target filter, i.e. labeled multi-Bernoulli (LMB) filter. The method has been developed for two practical applications of tracking a coordinated swarm and vehicles. The method has been tested for a complex vehicle tracking dataset and compared with the LMB filter through the OSPA and OSPA$^{(2)}$ metrics. The results demonstrate that the proposed interaction-aware method depicts considerable performance enhancement over the LMB filter in terms of the selected metrics.

Shima Rashidi, Ruwan Tennakoon, Aref Miri Rekavandi et al.

Distribution shift between train (source) and test (target) datasets is a common problem encountered in machine learning applications. One approach to resolve this issue is to use the Unsupervised Domain Adaptation (UDA) technique that carries out knowledge transfer from a label-rich source domain to an unlabeled target domain. Outliers that exist in either source or target datasets can introduce additional challenges when using UDA in practice. In this paper, $α$-divergence is used as a measure to minimize the discrepancy between the source and target distributions while inheriting robustness, adjustable with a single parameter $α$, as the prominent feature of this measure. Here, it is shown that the other well-known divergence-based UDA techniques can be derived as special cases of the proposed method. Furthermore, a theoretical upper bound is derived for the loss in the target domain in terms of the source loss and the initial $α$-divergence between the two domains. The robustness of the proposed method is validated through testing on several benchmarked datasets in open-set and partial UDA setups where extra classes existing in target and source datasets are considered as outliers.

WeiQin Chuah, Ruwan Tennakoon, Reza Hoseinnezhad et al.

Deep learning techniques often perform poorly in the presence of domain shift, where the test data follows a different distribution than the training data. The most practically desirable approach to address this issue is Single Domain Generalization (S-DG), which aims to train robust models using data from a single source. Prior work on S-DG has primarily focused on using data augmentation techniques to generate diverse training data. In this paper, we explore an alternative approach by investigating the robustness of linear operators, such as convolution and dense layers commonly used in deep learning. We propose a novel operator called XCNorm that computes the normalized cross-correlation between weights and an input feature patch. This approach is invariant to both affine shifts and changes in energy within a local feature patch and eliminates the need for commonly used non-linear activation functions. We show that deep neural networks composed of this operator are robust to common semantic distribution shifts. Furthermore, our empirical results on single-domain generalization benchmarks demonstrate that our proposed technique performs comparably to the state-of-the-art methods.

Xiaoying Wang, Reza Hoseinnezhad, Amirali K. Gostar et al.

This paper presents a novel statistical information fusion method to integrate multiple-view sensor data in multi-object tracking applications. The proposed method overcomes the drawbacks of the commonly used Generalized Covariance Intersection method, which considers constant weights allocated for sensors. Our method is based on enhancing the Generalized Covariance Intersection with adaptive weights that are automatically tuned based on the amount of information carried by the measurements from each sensor. To quantify information content, Cauchy-Schwarz divergence is used. Another distinguished characteristic of our method lies in the usage of the Labeled Multi-Bernoulli filter for multi-object tracking, in which the weight of each sensor can be separately adapted for each Bernoulli component of the filter. The results of numerical experiments show that our proposed method can successfully integrate information provided by multiple sensors with different fields of view. In such scenarios, our method significantly outperforms the state of art in terms of inclusion of all existing objects and tracking accuracy.

Amirali K. Gostar, Reza Hoseinnezhad, Alireza Bab-Hadiashar

A novel sensor control solution is presented, formulated within a Multi-Bernoulli-based multi-target tracking framework. The proposed method is especially designed for the general multi-target tracking case, where no prior knowledge of the clutter distribution or the probability of detection profile are available. In an information theoretic approach, our method makes use of Rènyi divergence as the reward function to be maximized for finding the optimal sensor control command at each step. We devise a Monte Carlo sampling method for computation of the reward. Simulation results demonstrate successful performance of the proposed method in a challenging scenario involving five targets maneuvering in a relatively uncertain space with unknown distance-dependent clutter rate and probability of detection.

Ruwan Tennakoon, Alireza Sadri, Reza Hoseinnezhad et al.

Identifying the underlying models in a set of data points contaminated by noise and outliers, leads to a highly complex multi-model fitting problem. This problem can be posed as a clustering problem by the projection of higher order affinities between data points into a graph, which can then be clustered using spectral clustering. Calculating all possible higher order affinities is computationally expensive. Hence in most cases only a subset is used. In this paper, we propose an effective sampling method to obtain a highly accurate approximation of the full graph required to solve multi-structural model fitting problems in computer vision. The proposed method is based on the observation that the usefulness of a graph for segmentation improves as the distribution of hypotheses (used to build the graph) approaches the distribution of actual parameters for the given data. In this paper, we approximate this actual parameter distribution using a k-th order statistics based cost function and the samples are generated using a greedy algorithm coupled with a data sub-sampling strategy. The experimental analysis shows that the proposed method is both accurate and computationally efficient compared to the state-of-the-art robust multi-model fitting techniques. The code is publicly available from https://github.com/RuwanT/model-fitting-cbs.

WeiQin Chuah, Ruwan Tennakoon, Reza Hoseinnezhad et al.

State-of-the-art stereo matching networks trained only on synthetic data often fail to generalize to more challenging real data domains. In this paper, we attempt to unfold an important factor that hinders the networks from generalizing across domains: through the lens of shortcut learning. We demonstrate that the learning of feature representations in stereo matching networks is heavily influenced by synthetic data artefacts (shortcut attributes). To mitigate this issue, we propose an Information-Theoretic Shortcut Avoidance~(ITSA) approach to automatically restrict shortcut-related information from being encoded into the feature representations. As a result, our proposed method learns robust and shortcut-invariant features by minimizing the sensitivity of latent features to input variations. To avoid the prohibitive computational cost of direct input sensitivity optimization, we propose an effective yet feasible algorithm to achieve robustness. We show that using this method, state-of-the-art stereo matching networks that are trained purely on synthetic data can effectively generalize to challenging and previously unseen real data scenarios. Importantly, the proposed method enhances the robustness of the synthetic trained networks to the point that they outperform their fine-tuned counterparts (on real data) for challenging out-of-domain stereo datasets.

Ammar Mansoor Kamoona, Amirali Khodadadian Gostar, Alireza Bab-Hadiashar et al.

In this paper, we propose an efficient approach for industrial defect detection that is modeled based on anomaly detection using point pattern data. Most recent works use \textit{global features} for feature extraction to summarize image content. However, global features are not robust against lighting and viewpoint changes and do not describe the image's geometrical information to be fully utilized in the manufacturing industry. To the best of our knowledge, we are the first to propose using transfer learning of local/point pattern features to overcome these limitations and capture geometrical information of the image regions. We model these local/point pattern features as a random finite set (RFS). In addition we propose RFS energy, in contrast to RFS likelihood as anomaly score. The similarity distribution of point pattern features of the normal sample has been modeled as a multivariate Gaussian. Parameters learning of the proposed RFS energy does not require any heavy computation. We evaluate the proposed approach on the MVTec AD dataset, a multi-object defect detection dataset. Experimental results show the outstanding performance of our proposed approach compared to the state-of-the-art methods, and the proposed RFS energy outperforms the state-of-the-art in the few shot learning settings.

Ayman Mukhaimar, Ruwan Tennakoon, Chow Yin Lai et al.

The task of learning from point cloud data is always challenging due to the often occurrence of noise and outliers in the data. Such data inaccuracies can significantly influence the performance of state-of-the-art deep learning networks and their ability to classify or segment objects. While there are some robust deep learning approaches, they are computationally too expensive for real-time applications. This paper proposes a deep learning solution that includes a novel robust pooling layer which greatly enhances network robustness and performs significantly faster than state-of-the-art approaches. The proposed pooling layer looks for data a mode/cluster using two methods, RANSAC, and histogram, as clusters are indicative of models. We tested the pooling layer into frameworks such as Point-based and graph-based neural networks, and the tests showed enhanced robustness as compared to robust state-of-the-art methods.

Ammar Mansoor Kamoona, Amirali Khodadadian Gostar, Alireza Bab-Hadiashar et al.

Defect detection in the manufacturing industry is of utmost importance for product quality inspection. Recently, optical defect detection has been investigated as an anomaly detection using different deep learning methods. However, the recent works do not explore the use of point pattern features, such as SIFT for anomaly detection using the recently developed set-based methods. In this paper, we present an evaluation of different point pattern feature detectors and descriptors for defect detection application. The evaluation is performed within the random finite set framework. Handcrafted point pattern features, such as SIFT as well as deep features are used in this evaluation. Random finite set-based defect detection is compared with state-of-the-arts anomaly detection methods. The results show that using point pattern features, such as SIFT as data points for random finite set-based anomaly detection achieves the most consistent defect detection accuracy on the MVTec-AD dataset.

WeiQin Chuah, Ruwan Tennakoon, Reza Hoseinnezhad et al.

Stereo vision generally involves the computation of pixel correspondences and estimation of disparities between rectified image pairs. In many applications, including simultaneous localization and mapping (SLAM) and 3D object detection, the disparities are primarily needed to calculate depth values and the accuracy of depth estimation is often more compelling than disparity estimation. The accuracy of disparity estimation, however, does not directly translate to the accuracy of depth estimation, especially for faraway objects. In the context of learning-based stereo systems, this is largely due to biases imposed by the choices of the disparity-based loss function and the training data. Consequently, the learning algorithms often produce unreliable depth estimates of foreground objects, particularly at large distances~($>50$m). To resolve this issue, we first analyze the effect of those biases and then propose a pair of novel depth-based loss functions for foreground and background, separately. These loss functions are tunable and can balance the inherent bias of the stereo learning algorithms. The efficacy of our solution is demonstrated by an extensive set of experiments, which are benchmarked against state of the art. We show on KITTI~2015 benchmark that our proposed solution yields substantial improvements in disparity and depth estimation, particularly for objects located at distances beyond 50 meters, outperforming the previous state of the art by $10\%$.

Ayman Mukhaimar, Ruwan Tennakoon, Chow Yin Lai et al.

In this paper, we present a robust spherical harmonics approach for the classification of point cloud-based objects. Spherical harmonics have been used for classification over the years, with several frameworks existing in the literature. These approaches use variety of spherical harmonics based descriptors to classify objects. We first investigated these frameworks robustness against data augmentation, such as outliers and noise, as it has not been studied before. Then we propose a spherical convolution neural network framework for robust object classification. The proposed framework uses the voxel grid of concentric spheres to learn features over the unit ball. Our proposed model learn features that are less sensitive to data augmentation due to the selected sampling strategy and the designed convolution operation. We tested our proposed model against several types of data augmentation, such as noise and outliers. Our results show that the proposed model outperforms the state of art networks in terms of robustness to data augmentation.

Ammar Mansoor Kamoona, Amirali Khodadadian Gosta, Alireza Bab-Hadiashar et al.

In this paper, we propose a weakly supervised deep temporal encoding-decoding solution for anomaly detection in surveillance videos using multiple instance learning. The proposed approach uses both abnormal and normal video clips during the training phase which is developed in the multiple instance framework where we treat video as a bag and video clips as instances in the bag. Our main contribution lies in the proposed novel approach to consider temporal relations between video instances. We deal with video instances (clips) as a sequential visual data rather than independent instances. We employ a deep temporal and encoder network that is designed to capture spatial-temporal evolution of video instances over time. We also propose a new loss function that is smoother than similar loss functions recently presented in the computer vision literature, and therefore; enjoys faster convergence and improved tolerance to local minima during the training phase. The proposed temporal encoding-decoding approach with modified loss is benchmarked against the state-of-the-art in simulation studies. The results show that the proposed method performs similar to or better than the state-of-the-art solutions for anomaly detection in video surveillance applications.

Tharindu Rathnayake, Reza Hoseinnezhad, Ruwan Tennakoon et al.

This paper presents a track-before-detect labeled multi-Bernoulli filter tailored for industrial mobile platform safety applications. We derive two application specific separable likelihood functions that capture the geometric shape and colour information of the human targets who are wearing a high visible vest. These likelihoods are then used in a labeled multi-Bernoulli filter with a novel two step Bayesian update. Preliminary simulation results show that the proposed solution can successfully track human workers wearing a luminous yellow colour vest in an industrial environment.

Amirali K. Gostar, Reza Hoseinnezhad, Alireza Bab-Hadiashar

This paper presents a sensor-control method for choosing the best next state of the sensor(s), that provide(s) accurate estimation results in a multi-target tracking application. The proposed solution is formulated for a multi-Bernoulli filter and works via minimization of a new estimation error-based cost function. Simulation results demonstrate that the proposed method can outperform the state-of-the-art methods in terms of computation time and robustness to clutter while delivering similar accuracy.