HSHasan Sajid

Hasan Sajid

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Hasan Sajid has a total of 33 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    A scalable deep attention mechanism of instance segmentation for the investigation of chromosome

    by Neelam Umbreen, Sara Ali, Hasan Sajid, Yasar Ayaz, Shrooq Alsenan, Yunyoung Nam, So Yeon Kim, Muhammad Baber Sial


  • MSE-GCN: A Multiscale Spatiotemporal Feature Aggregation Enhanced Efficient Graph Convolutional Network for Dynamic Sign Language Recognition

    by Neelma Naz, Hasan Sajid, Sara Ali, Osman Hasan, Muhammad Khurram Ehsan

    Graph convolution networks have emerged as an active area of research for skeleton-based sign language recognition (SLR). One essential problem in this approach is to efficiently extract the most discriminative features capable of modeling short-range and long-range spatial and temporal information over all skeleton joints while ensuring low inference costs. To address this issue, we propose a novel multi-scale efficient graph convolutional network (MSE-GCN) for skeleton-based SLR. The proposed network makes use of separable convolution layers set in a multi-scale setting and embedded in a multi branch (MB) network along with an early fusion scheme, resulting in an accurate, computationally efficient, and faster system. In addition, we have proposed a novel hybrid attention module, named Spatial Temporal Joint Part attention (ST-JPA) to distinguish the most important body parts as well as most informative joints in the specific frames from the whole sign sequence. The performance of proposed network (MSE-GCN) is evaluated on five challenging sign language datasets, WLASL-100, WLASL-300, WLASL-1000, MINDS-Libras, and LIBRAS-UFOP achieving state-of-the-art (SOTA) accuracies of 85.27%, 81.59%, 71.75%, 97.442 ± 1.01%, and 88.59±3.60%, respectively while incurring lower computational costs.


  • Open Access

    Fine-Tuning Quadcopter Control Parameters via Deep Actor-Critic Learning Framework: An Exploration of Nonlinear Stability Analysis and Intelligent Gain Tuning

    by Hassan Moin, Umer Hameed Shah, Muhammad Jawad Khan, Hasan Sajid

    Quadcopters have underactuated, nonlinear, and coupled dynamics, making their control a challenging endeavor. However, PID controllers have exhibited remarkable performance for such systems in a variety of circumstances, including obstacle avoidance, trajectory tracking, and route planning. The outer loop handles mission-level goals, while the inner loop is responsible for its stability and control. Nevertheless, gain optimization for nonlinear systems using heuristics or rule-based approaches is a laborious, time-consuming, and challenging task. This study implements an optimal gain self-tuning framework for the altitude, attitude, and position controllers of a 6 DoF nonlinear drone system using a Deep Reinforcement Learning (DRL) algorithm with continuous observation and action spaces. The state equations are determined using Lagrange’s (energy-based) technique, whilst the aerodynamic coefficients are computed numerically using Blade Element Momentum Theory (BEMT). In addition, the asymptotic stability of the cascaded closed loop nonlinear system is investigated using the theory of Lyapunov. The proposed technique is validated via simulations, which demonstrate that the quadcopter can closely follow a specified trajectory by using the optimized gains. Most notably, these optimal gains fulfill the constraints derived from Lyapunov’s stability analysis, suggesting that the RL technique is an incredibly powerful instrument that can evaluate the uncertainties present in any complicated nonlinear system.


  • Open Access

    Stain Normalization of Hematology Slides using Neural Color Transfer

    by M. Muneeb Arshad, Hasan Sajid, M. Jawad Khan

    Deep learning is popularly used for analyzing pathology images, but variations in image properties can limit the effectiveness of the models. The study aims to develop a method that transfers the variability present in the training set to unseen images, improving the model's ability to make accurate inferences. YOLOv5 was trained on peripheral blood and bone marrow sample images and Neural Color Transfer techniques were used to incorporate invariance. The results showed significant improvement in detecting WBCs from untrained samples after normalization, highlighting the potential of deep learning-based normalization techniques for inference robustness.


  • Top Cited Papers
    Open Access

    Signgraph: An Efficient and Accurate Pose-Based Graph Convolution Approach Toward Sign Language Recognition

    by Neelma Naz, Hasan Sajid, Sara Ali, Osman Hasan, Muhammad Khurram Ehsan

    Sign language recognition (SLR) enables the deaf and speech-impaired community to integrate and communicate effectively with the rest of society. Word level or isolated SLR is a fundamental yet complex task with the main objective of using models to correctly recognize signed words. Sign language consists of very fast and complex hand, body, face movements, and mouthing cues that make the task very challenging. Several input modalities; RGB, optical Flow, RGB-D, and pose/skeleton have been proposed for SLR. However, the complexity of these modalities and the state-of-the-art (SOTA) methodologies tend to be exceedingly sophisticated and over-parametrized. In this paper, our focus is to use the hands and body poses as an input modality. One major problem in pose-based SLR is extracting the most valuable and distinctive features for all skeleton joints. In this regard, we propose an accurate, efficient, and lightweight pose-based pipeline leveraging a graph convolution network (GCN) along with residual connections and a bottleneck structure. The proposed architecture not only facilitates efficient learning during model training providing significantly improved accuracy scores but also alleviates computational complexity. With the proposed architecture in place, we are able to achieve improved accuracies on three different subsets of the WLASL dataset and the LSA-64 dataset. Our proposed model outperforms previous SOTA pose-based methods by providing a relative improvement of 8.91%, 27.62%, and 26.97% for WLASL-100, WLASL-300, and WLASL-1000 subsets. Moreover, our proposed model also outperforms previous SOTA appearance-based methods by providing a relative improvement of 2.65% and 5.15% for WLASL-300 and WLASL-1000 subsets. For the LSA-64 dataset, our model is able to achieve 100% test recognition accuracy. We are able to achieve this improved performance with far less computational cost as compared to existing appearance-based methods.


  • Open Access

    m2caiSeg: Semantic Segmentation of Laparoscopic Images using Convolutional Neural Networks

    by Salman Maqbool, Aqsa Riaz, Hasan Sajid, Osman Hasan

    Autonomous surgical procedures, in particular minimal invasive surgeries, are the next frontier for Artificial Intelligence research. However, the existing challenges include precise identification of the human anatomy and the surgical settings, and modeling the environment for training of an autonomous agent. To address the identification of human anatomy and the surgical settings, we propose a deep learning based semantic segmentation algorithm to identify and label the tissues and organs in the endoscopic video feed of the human torso region. We present an annotated dataset, m2caiSeg, created from endoscopic video feeds of real-world surgical procedures. Overall, the data consists of 307 images, each of which is annotated for the organs and different surgical instruments present in the scene. We propose and train a deep convolutional neural network for the semantic segmentation task. To cater for the low quantity of annotated data, we use unsupervised pre-training and data augmentation. The trained model is evaluated on an independent test set of the proposed dataset. We obtained a F1 score of 0.33 while using all the labeled categories for the semantic segmentation task. Secondly, we labeled all instruments into an 'Instruments' superclass to evaluate the model's performance on discerning the various organs and obtained a F1 score of 0.57. We propose a new dataset and a deep learning method for pixel level identification of various organs and instruments in a endoscopic surgical scene. Surgical scene understanding is one of the first steps towards automating surgical procedures.


  • Open Access

    Secure Route-Obfuscation Mechanism with Information-Theoretic Security for Internet of Things

    by Abid Rauf, Zhaohong Wang, Hasan Sajid, Muhammad Ali Tahir

    As accessibility of networked devices becomes more and more ubiquitous, groundbreaking applications of the Internet of Things (IoT) find their place in many aspects of our society. The exploitation of these devices is the main reason for the cyberattacks in IoT networks. Security design is still an open problem and a crucial step in making IoT applications successful. In dicey environments, such as e-health, smart grid, and smart cities, real-time commands must reach the end devices in the scale of milliseconds. Traditional public-key cryptosystem, albeit necessary in the context of general Internet security, falls short in establishing new session keys in the scale of milliseconds for critical messages. In this paper, a systematic perspective for securing IoT communication, specifically satisfying the real-time constraint against certain adversaries in realistic settings. First, at the network layer, we propose a secret random route computation scheme using the software-defined network (SDN) based on a capability scheme using the network actions. The computed routes are random in the eyes of the eavesdropper. Second, at the application layer, the source breaks command messages into secret shares and sends them through the network to the destination. Only the legitimate destination device can reconstruct the command. The secret sharing scheme is efficient compared to PKI and comes with information-theoretic security against adversaries. Our proof formalizes the notion of security of the proposed scheme, and our simulations validate our design.


  • Iterating Tensor Voting: A Perceptual Grouping Approach for Crack Detection on EL Images

    by Kun Liu, Haowei Yan, Kai Meng, Haiyong Chen, Hasan Sajid

    The surface of a multicrystal solar cell shows multiple crystal grains of random shapes and sizes. It creates an inhomogeneous texture in the surface, which brings great difficulty to automatic crack detection of polycrystalline solar surface. As a perceptual grouping approach, tensor voting can extract curvilinear structures such as lines and curves from noisy, binary data in 2-D or 3-D, without invoking specific object or model. However, traditional tensor voting can be susceptible to the gap problem and structural noise. To address the problems mentioned above, a new iterative tensor voting algorithm is presented, which efficacy bases on iterative refinements of the curvilinear structures. By combining the proximity and continuity of Gestalt principles, in each iteration step, a new decay function is redefined according to the difference of angle between the voter and receiver to rebuild the voting field, which makes the points that lie on curvilinear structures vote more information (a bigger tensor) to the ones with the same attribute. The proposed method can solve the gap problem and is robust to structural noises. The experimental results show that the proposed method can detect crack on the inhomogeneous textured surface and achieve an average detection rate of 95.2% on the industrial data set.


  • ZoomCount: A Zooming Mechanism for Crowd Counting in Static Images

    by Usman Sajid, Hasan Sajid, Hongcheng Wang, Guanghui Wang

    This paper proposes a novel approach for crowd counting in low to high density scenarios in static images. Current approaches cannot handle huge crowd diversity well and thus perform poorly in extreme cases, where the crowd density in different regions of an image is either too low or too high, leading to crowd underestimation or overestimation. The proposed solution is based on the observation that detecting and handling such extreme cases in a specialized way leads to better crowd estimation. Additionally, existing methods find it hard to differentiate between the actual crowd and the cluttered background regions, resulting in further count overestimation. To address these issues, we propose a simple yet effective modular approach, where an input image is first subdivided into fixed-size patches and then fed to a four-way classification module labeling each image patch as low, medium, high-dense or no-crowd. This module also provides a count for each label, which is then analyzed via a specifically devised novel decision module to decide whether the image belongs to any of the two extreme cases (very low or very high density) or a normal case. Images, specified as high- or low-density extreme or a normal case, pass through dedicated zooming or normal patch-making blocks respectively before routing to the regressor in the form of fixed-size patches for crowd estimate. Extensive experimental evaluations demonstrate that the proposed approach outperforms the state-of-the-art methods on four benchmarks under most of the evaluation criteria.