Print Email Facebook Twitter Binary Neural Networks for Object Detection Title Binary Neural Networks for Object Detection Author Wang, Yizhou (TU Delft Electrical Engineering, Mathematics and Computer Science) Contributor Al-Ars, Z. (mentor) Pan, W. (graduation committee) van Genderen, A.J. (graduation committee) Zhu, B. (graduation committee) Degree granting institution Delft University of Technology Programme Electrical Engineering | Embedded Systems Date 2019-08-29 Abstract In the past few years, convolutional neural networks (CNNs) have been widely utilized and shown state-of-the-art performances on computer vision tasks. However, CNN based approaches usually require a large amount of storage, run-time memory, as well as computation power in both training and inference time, which are usually used on GPU based machines to ensure the speed for inferences. But they are usually insufficient to be deployed on low-power applications. Although many approaches were proposed to compress and accelerate the CNN models, most of them were only evaluated on relatively simple problems (e.g. image classification), which only support limited real-world applications. Especially, among those methods, binary quantization can achieve very high model compression, but only a few works have been observed to utilize it on more complex tasks. Therefore, the exploration and evaluations of applying binary quantization on more complex tasks like object detection are worthwhile, which can be used in much more applications like autonomous driving and face detection. In this project, we apply and evaluate two different binary quantization approaches, named ABC-Net and PA-Net on object detection tasks. Also, we specify the exact implementation details for the binary convolutional operations in this project. As a result, we can achieve maximally 6.1× (around 16% of the full-precision model) compression, and minimal 2.5% accuracy reduction for weight quantization. The weight quantized models were able to outperform some existing real-time detectors in terms of both accuracy and storage size. Although large accuracy reduction was observed for input quantization, the quantized model could still maintain an acceptable accuracy compared to existing real-time object detectors. Subject Neural Network QuantizationDeep LearningObject DetectionComputer VisionArtificial Intelligence To reference this document use: http://resolver.tudelft.nl/uuid:9f0da106-82ea-4f2e-9cd5-8bc834885d6f Embargo date 2020-09-01 Part of collection Student theses Document type master thesis Rights © 2019 Yizhou Wang Files PDF MSc_Thesis_Yizhou_Wang.pdf 1.8 MB Close viewer /islandora/object/uuid:9f0da106-82ea-4f2e-9cd5-8bc834885d6f/datastream/OBJ/view