Browsing by Author "Prasad, S. P. Kavyashree"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Compressed MobileNet V3: An efficient CNN for resource constrained platforms
    (2021-05) Prasad, S. P. Kavyashree; El-Sharkawy, Mohamed; King, Brian; Rizkalla, Maher
    Computer Vision is a mathematical tool formulated to extend human vision to machines. This tool can perform various tasks such as object classification, object tracking, motion estimation, and image segmentation. These tasks find their use in many applications, namely robotics, self-driving cars, augmented reality, and mobile applications. However, opposed to the traditional technique of incorporating handcrafted features to understand images, convolution neural networks are being used to perform the same function. Computer vision applications widely use CNNs due to their stellar performance in interpreting images. Over the years, there have been numerous advancements in machine learning, particularly to CNNs.However, the need to improve their accuracy, model size and complexity increased, making their deployment in restricted environments a challenge. Many researchers proposed techniques to reduce the size of CNN while still retaining its accuracy. Few of these include network quantization, pruning, low rank, and sparse decomposition and knowledge distillation. Some methods developed efficient models from scratch. This thesis achieves a similar goal using design space exploration techniques on the latest variant of MobileNets, MobileNet V3. Using DPD blocks, escalation in the number of expansion filters in some layers and mish activation function MobileNet V3 is reduced to 84.96% in size and made 0.2% more accurate. Furthermore, it is deployed in NXP i.MX RT1060 for image classification on CIFAR-10 dataset.
  • Thumbnail Image
    Item
    Deployment of Compressed MobileNet V3 on iMX RT 1060
    (IEEE Xplore, 2021-04) Prasad, S. P. Kavyashree; El-Sharkawy, Mohamed; Electrical and Computer Engineering, School of Engineering and Technology
    Deep Neural Networks (DNN) are prominent in most applications today. From self-driving cars, sentiment analysis, surveillance systems, and robotics, they have been used extensively. Among DNNs, Convolutional Neural Networks (CNN) have achieved massive success in computer vision applications as the human visual system inspires their architecture. However, striving to achieve higher accuracies, CNN complexity, parameters, and layers were increased, which led to a drastic surge in their size, making their deployment challenging. Over the years, many researchers have proposed various techniques to alleviate this issue-one of them being Design Space Exploration (DSE) to minimize size and computation with little compromise to accuracy. MobileNet V3 is one such architecture designed to achieve good accuracy while being mindful of resources. It produces an accuracy of 88.93% on CIFAR-10 with a size of 15.3MB. This paper further reduces its size to 2.3MB while boosting its accuracy to 89.13% using DSE techniques. It is then deployed into NXP's i.MX RT1060 Advanced Driver Assistance System (ADAS) platform.