Deep learning-based classification and detection algorithms have emerged as a powerful tool for vehicle detection in intelligent transportation systems. The limitations of the number of high-quality labeled training samples makes the single vehicle detection methods incapable of accomplishing acceptable accuracy in road vehicle detection. This paper presents detection and classification of vehicles on publicly available datasets by utilizing the YOLO-v5 architecture. This paper’s findings utilize the concept of transfer learning through fine tuning the weights of the pre-trained YOLO-v5 architecture. To employ the concept of transfer learning, extensive data sets of images and videos of the congested traffic patterns were collected by the authors. These datasets were made more comprehensive by pointing various attributes, for instance high- and low-density traffic patterns, occlusions, and different weather circumstances. All of these gathered datasets were manually annotated. Ultimately, the improved YOLO-v5 structure becomes accustomed to any difficult traffic patterns. By fine-tuning the pre-trained network through our datasets, our proposed YOLO-v5 has exceeded several other traditional vehicle detection methods in terms of detection accuracy and execution time. Detailed simulations performed on the PKU, COCO, and DAWN datasets demonstrate the effectiveness of the proposed method in various challenging situations.

The diversion of a driver’s attention from driving can be catastrophic. Given that conventional button- and touch-based interfaces may distract the driver, developing novel distraction-free interfaces for the various devices present in cars has becomes necessary. Hand gesture recognition may provide an alternative interface inside cars. Given that cars are the targeted application area, we determined the optimal location for the radar sensor, so that the signal reflected from the driver’s hand during gesturing is unaffected by interference from the motion of the driver’s body or other motions within the car. We implemented a Convolutional Neural Network-based technique to recognize the finger-counting-based hand gestures using an Impulse Radio (IR) radar sensor. The accuracy of the proposed method was sufficiently high for real-world applications.

A compression technique for still digital images is proposed with deep neural networks (DNNs) employing rectified linear units (ReLUs). We tend to exploit the DNNs capabilities to find a reasonable estimate of the underlying compression/decompression relationships. We aim for a DNN for image compression purpose that has better generalization property and reduced training time and support real time operation. The use of ReLUs which map more plausibly to biological neurons, makes the training of our DNN significantly faster, shortens the encoding/decoding time, and improves its generalization ability. The introduction of the ReLUs establishes an efficient gradient propagation, induces sparsity in the proposed network, and is efficient in terms of computations making these networks suitable for real time compression systems. Experiments performed on standard real world images show that using ReLUs instead of logistic sigmoid units speeds up the training of the DNN by converging markedly faster. The evaluation of objective and subjective quality of reconstructed images also proves that our DNN achieves better generalization as most of the images are never seen by the network before.

The paper presents a novel methodology based on machine learning to optimize medical benefits in healthcare settings, i.e., corporate, private, public or statutory. The optimization is applied to design healthcare insurance packages based on the employee healthcare record. Moreover, with the advancement in the insurance industry, it is rapidly adapting mathematical and machine learning models to enhance insurance services like funds prediction, customer management and get better revenue from their businesses. However, conventional computing insurance packages and premium methods are time-consuming, designation specific, and not cost-effective. During the design of insurance packages, an employee’s needs should be given more importance than his/her designation or position in an organization. The design of insurance packages in healthcare is a non-trivial task due to the employees’ changing healthcare needs; therefore, using the proposed technique employees can be moved from their existing package to another depending upon his/her need. This provides the motivation to propose a methodology in which we applied machine learning concepts for designing need-based health insurance packages rather than professional tagging. By the design of need-based packages, medical benefit optimization which is the core goal of our proposed methodology is effectively achieved. Our proposed methodology derives insurance packages that are need-based and optimal based on our defined criteria. We achieved this by first applying the clustering technique to historical medical records. Subsequently, medical benefit optimization is achieved from these packages by applying a probability distribution model on five years employees’ insurance records. The designed technique is validated on real employees’ insurance records from a large enterprise.The proposed design provides 25% optimization on medical benefit amount compared to current medical benefits amount therefore, gives better healthcare to all the employees.

Introduction: Convolutional neural networks (CNNs) have maintained their dominance in deep learning methods for human action recognition (HAR) and other computer vision tasks. However, the need for a large amount of training data always restricts the performance of CNNs. Method: This paper is inspired by the two-stream network, where a CNN is deployed to train the network by using the spatial and temporal aspects of an activity, thus exploiting the strengths of both networks to achieve better accuracy. Contributions: Our contribution is twofold: first, we deploy an enhanced spatial stream, and it is demonstrated that models pre-trained on a larger dataset, when used in the spatial stream, yield good performance instead of training the entire model from scratch. Second, a dataset augmentation technique is presented to minimize overfitting of CNNs, where we increase the dataset size by performing various transformations on the images such as rotation and flipping, etc. Results: UCF101 is a standard benchmark dataset for action videos, and our architecture has been trained and validated on it. Compared with the other two-stream networks, our results outperformed them in terms of accuracy.