Sharif Digital Repository

The objective of OCR systems is to retrieve machine-encoded text from a raster image. Despite the abundance of powerful OCR algorithms for English, there are not many for Farsi. Our proposed algorithm is comprised of pre-processing, line detection, sub-word detection and segmentation, feature extraction and classification. Furthermore, hardware implementation and acceleration of this system on a GPGPU is presented. This algorithm was tested on 5 fonts including Titr, Lotus,Yekan, Koodak and Nazanin and an average accuracy above 90% was achieved  

Spiking Neural Networks(SNN) are networks which are consisted of layers of neurons, like other typical artificial neural networks. The main difference between SNN and other neural networks is the type of data transportation among neurons which is done by spikes. Spiking neural networks and their models are considered as the nearest networks and neurons to animals’ nervous systems. In aspects of hardware implementation, the type of data transportation in SNN causes them to be ultra-low power. So, implementation of these networks on chips like FPGA and also usage of SNN in applications with high processing load have startling germination, recently. In this work, we have tried to propose some... 

There have been many works on generative networks and image generation in the past few years, but the problem with this work is that there is no control over the generated images. The goal of disentangled image synthesis is to generate new images with specific detail and have control over the generated images. Image-based virtual try-on aims to synthesize the customer image with an in-shop clothes image to acquire seamless and natural try-on results, which have attracted increasing attention. The main procedures of image-based virtual try-on usually consist of clothes image generation and try-on image synthesis. In contrast, prior arts cannot guarantee satisfying clothes results when facing... 

Many mobile applications running on smartphones, wearable devices, tiny autonomous robots and IoT devices would potentially benefit from the accuracy and scalability of deep CNN-based machine learning algorithms. However,performance and energy consumption limitations make the execution of such computationally intensive algorithms on embedded mobile devices prohibitive.We present a GPU-accelerated engine, dubbed mCNN, for execution of trained deep CNNs on mobile platforms. The proposed solution takes the trained model as input and automatically optimizes its parallel implementation on the target mobile platform for efficient use of hardware resources such as mobile GPU threads and SIMD units.... 

Many chip manufacturers have recently introduced high-performance deep-learning hardware accelerators. In modern GPUs, programmable tensor cores accelerate the heavy operations involved in deep neural networks. This paper presents a novel solution to re-purpose tensor cores in modern GPUs for high-throughput implementation of turbo decoders. Turbo codes closely approach Shannon’s limit on channel capacity, and are widely used in many state-of-the-art wireless systems including satellite communications and mobile communications. Experimental evaluations show that the proposed solution achieves about 1.2 Gbps throughput, which is higher compared to previous GPU-accelerated solutions  

In the current era, artificial intelligence and machine learning have become vital and widely used technologies across various industries. These technologies enable companies and organizations to optimize processes, predict trends, and uncover hidden patterns in data with high accuracy and speed. However, fully leveraging the capabilities of AI and ML requires the effective and efficient deployment of ML models in production environments. MLOps, a combination of DevOps and ML concepts, aids in managing the lifecycle of ML models from development to deployment and maintenance. In this research, due to international sanctions and limited access to external services such as Google Vertex AI,... 

In this project, a method is emoloyed to implement a Viterbi decoder on GPGPU. This method is based on combining all steps of the algorithm. This combination has some challenges that are related to differences between different steps of the algorithm. So in this project, some solutions are found to handle these challenges and a high-throughput Viterbi decoder is acheived  

Cardiovascular diseases are the leading cause of death globally so early diagnosis of them is important. Many researchers focused on this field. First signs of cardiac diseases appear in the electrocardiogram signal. This signal represents the electrical activity of the heart so it’s primarily used for the detection and classification of cardiac arrhythmias. Permanent monitoring of this signal is not possible for specialists so we should do this by means of Artificial Intelligence. In this thesis, we use recurrent neural networks to classify electrocardiogram’s arrhythmias. This deep learning method, use two sources of data to learn from. The first part of data is global for everyone and the...