Chirine Riachy

Bio: Chirine Riachy is an academic researcher from Northumbria University. The author has contributed to research in topics: Feature extraction & Feature (computer vision). The author has an hindex of 3, co-authored 5 publications receiving 19 citations.

Video-based Person Re-Identification Using Unsupervised Tracklet Matching

Abstract: Despite the significant improvement in accuracy supervised learning has brought into person re-identification (re-id), the availability of sufficient fully annotated data from concerned camera-views poses a problem for real-life applications. To alleviate the burden of intensive data annotation, one way is to resort to unsupervised methods. This has motivated us to propose a novel algorithm for unsupervised video-based person re-id applications. To achieve this, the frames of a person video tracklet are divided into a set of clusters that are subsequently matched using a distance measure based on the Naive Bayes nearest neighbor algorithm and Spearman distance. Knowing that person's sequences may suffer from substantial changes in viewpoint, pose, and illumination distortions, our technique allows the rejection of poor and noisy clusters while retaining the most discriminative ones for matching. Experiments on three widely used datasets for video person re-id PRID2011, iLIDS-VID and MARS have been carried out, and the results demonstrate the superiority of the proposed approach.

Online Screening of X-System Music Playlists Using an Integrative Wellbeing Model Informed by the Theory of Autopoiesis

Abstract: Research suggests that music has a powerful effect on the human mind and body. This article explores the impact of music as an intervention. For this purpose, the X-System technology is used to curate relaxing and enlivening music playlists designed to positively impact wellbeing and emotional state during the COVID-19 pandemic. A wellbeing model grounded in autopoietic theory of self-organisation in living systems is developed to inform the evaluation of the impact of the intervention and ensure the reliability of the data. More specifically, data quality is enhanced by focusing the participants’ awareness on their immediate embodied experience of physical, emotional and relational wellbeing and sense of pleasure/displeasure prior to and after listening to a preferred playlist. The statistical analysis shows significant positive changes in emotional wellbeing, valence and sense of meaning ( $p ) with a medium effect size. It also reveals a statistically significant change for physical wellbeing ( $p=0.009$ ) with a small effect size. With the relaxing playlists leading to decrease in arousal levels and the enlivening playlists to an increase in activation, it is also concluded that appropriately curated playlists may be able to lead the listener to positive relaxation or activation states or indeed to positive mood change that may have health benefits.

Triplet Loss with Channel Attention for Person Re-identification

Abstract: The triplet loss function has seen extensive use within person re-identification. Most works focus on either improving the mining algorithm or adding new terms to the loss function itself. Our work instead concentrates on two other core components of the triplet loss that have been under-researched. First, we improve the standard Euclidean distance with dynamic weights, which are selected based on the standard deviation of features across the batch. Second, we exploit channel attention via a squeeze and excitation unit in the backbone model to emphasise important features throughout all layers of the model. This ensures that the output feature vector is a better representation of the image, and is also more suitable to use within our dynamically weighted Euclidean distance function. We demonstrate that our alterations provide significant performance improvement across popular reidentification data sets, including almost 10% mAP improvement on the CUHK03 data set. The proposed model attains results competitive with many state-of-the-art person re-identification models.

Person re-identification: Attribute-based feature evaluation

Abstract: Matching people across various camera views at different times and locations is called person re-identification (re-id). Despite remarkable recent advances in the field, the problem is still particularly challenging due to significant viewpoint angle variations, illumination changes, background clutter, occlusions, motion blur and pose variations. Although these challenges are widely acknowledged within the re-id community, no previous work has attempted to analyze the extent to which they individually contribute to performance deterioration, and whether current feature representations have succeeded to mitigate their effect. To address this matter, 4 publically available single-shot datasets were fully and manually annotated for these attributes. Subsequently, 6 state-of-the-art feature representations were evaluated considering each attribute separately. Extensive experimentation using XQDA distance metric has clearly shown that illumination variation remains the major problem hindering accurate re-id. The results also provide insights towards the design of new feature representations to improve the performance.

3D Gaussian descriptor for video-based person re-identification

Abstract: Despite being often considered less challenging than image-based person re-identification (re-id), video-based person re-id is still appealing as it mimics a more realistic scenario owing to the availability of pedestrian sequences from surveillance cameras. In order to exploit the temporal information provided, a number of feature extraction methods have been proposed. Although the features could be equally learned at a significantly higher computational cost, the scarce nature of labelled re-id datasets encourages the development of robust hand-crafted feature representations as an efficient alternative, especially when novel distance metrics or multi-shot ranking algorithms are to be validated. This paper presents a novel hand-crafted feature representation for video-based person re-id based on a 3-dimensional hierarchical Gaussian descriptor. Compared to similar approaches, the proposed descriptor (i) does not require any walking cycle extraction, hence avoiding the complexity of this task, (ii) can be easily fed into off-shelf learned distance metrics, (iii) and consistently achieves superior performance regardless of the matching method adopted. The performance of the proposed method was validated on PRID2011 and iLIDS-VID datasets outperforming similar methods on both benchmarks.

"Awareness of subtle emotional feelings: A comparison of long-term meditators and nonmeditators": Correction to Nielsen and Kaszniak (2006).

Abstract: The authors explored whether meditation training to enhance emotional awareness improves discrimination of subtle emotional feelings hypothesized to guide decision-making. Long-term meditators and nonmeditators were compared on measures of self-reported valence and arousal, skin conductance response (SCR), and facial electromyography (EMG) to masked and nonmasked emotional pictures, and on measures of heartbeat detection and self-reported emotional awareness. Groups responded similarly to nonmasked pictures. In the masked condition, only controls showed discrimination in valence self-reports. However, meditators reported greater emotional clarity than controls, and meditators with higher clarity had reduced arousal and improved valence discrimination in the masked condition. These findings provide qualified support for the somatic marker hypothesis and suggest that meditation may influence how emotionally ambiguous information is processed, regulated, and represented in conscious awareness.

Gait recognition for person re-identification

Abstract: Person re-identification across multiple cameras is an essential task in computer vision applications, particularly tracking the same person in different scenes. Gait recognition, which is the recognition based on the walking style, is mostly used for this purpose due to that human gait has unique characteristics that allow recognizing a person from a distance. However, human recognition via gait technique could be limited with the position of captured images or videos. Hence, this paper proposes a gait recognition approach for person re-identification. The proposed approach starts with estimating the angle of the gait first, and this is then followed with the recognition process, which is performed using convolutional neural networks. Herein, multitask convolutional neural network models and extracted gait energy images (GEIs) are used to estimate the angle and recognize the gait. GEIs are extracted by first detecting the moving objects, using background subtraction techniques. Training and testing phases are applied to the following three recognized datasets: CASIA-(B), OU-ISIR, and OU-MVLP. The proposed method is evaluated for background modeling using the Scene Background Modeling and Initialization (SBI) dataset. The proposed gait recognition method showed an accuracy of more than 98% for almost all datasets. Results of the proposed approach showed higher accuracy compared to obtained results of other methods result for CASIA-(B) and OU-MVLP and form the best results for the OU-ISIR dataset.

Exploiting Global Camera Network Constraints for Unsupervised Video Person Re-Identification

Abstract: Many unsupervised approaches have been proposed recently for the video-based re-identification problem since annotations of samples across cameras are time-consuming. However, higher-order relationships across the entire camera network are ignored by these methods, leading to contradictory outputs when matching results from different camera pairs are combined. In this paper, we address the problem of unsupervised video-based re-identification by proposing a consistent cross-view matching (CCM) framework, in which global camera network constraints are exploited to guarantee the matched pairs are with consistency. Specifically, we first propose to utilize the first neighbor of each sample to discover relations among samples and find the groups in each camera. Additionally, a cross-view matching strategy followed by global camera network constraints is proposed to explore the matching relationships across the entire camera network. Finally, we learn metric models for camera pairs progressively by alternatively mining consistent cross-view matching pairs and updating metric models using these obtained matches. Rigorous experiments on two widely-used benchmarks for video re-identification demonstrate the superiority of the proposed method over current state-of-the-art unsupervised methods; for example, on the MARS dataset, our method achieves an improvement of 4.2% over unsupervised methods, and even 2.5% over one-shot supervision-based methods for rank-1 accuracy.

Off-line Handwritten Signature Verification System: Artificial Neural Network Approach

Abstract: Nowadays, it is evident that signature is commonly used for personal verification, this justifies the necessity for an Automatic Verification System (AVS). Based on the application, verification could either be achieved Offline or Online. An online system uses the signature’s dynamic information; such information is captured at the instant the signature is generated. An offline system, on the other hand, uses an image (the signature is scanned). In this paper, some set of simple shaped geometric features are used in achieving offline Verification of signatures. These features include Baseline Slant Angle (BSA), Aspect Ratio (AR), and Normalized Area (NA), Center of Gravity as well as the line’s Slope that joins the Center of Gravities of the signature’s image two splits. Before the features extraction, a signature preprocessing is necessary to segregate its parts as well as to eliminate any available spurious noise. Primarily, System training is achieved via a signature record which was acquired from personalities whose signatures had to be validated through the system. An average signature is acquired for each subject as a result of incorporating the aforementioned features which were derived from a sample set of the subject’s true signatures. Therefore, a signature functions as the prototype for authentication against a requested test signature. The similarity measure within the feature space between the two signatures is determined by Euclidian distance. If the Euclidian distance is lower than a set threshold (i.e. analogous to the minimum acceptable degree of similarity), the test signature is certified as that of the claiming subject otherwise detected as a forgery. Details on the stated features, pre-processing, implementation, and the results are presented in this work.