About

Education

Research Interests

Research Experience

• Undergraduate Research Assistant, University of Tehran

  Feb, 2021 - Present

  • Working on Texture Synthesis problem in Single-View 3D Reconstruction

  • Best Undergraduate Project Award

  • Under the Supervision of Prof. Hadi Moradi and Prof. Reshad Hosseini

• Research and Development Engineer, Jadooye Fekr (JCo)

  Jan, 2019 - Present

  • Associated with machine learning, computer vision, time-series analysis, cryptocurrencies and python skills

Publications

Honors & Awards

Volunteer Work

Certificates

Projects Experience

• A Framework for Executing Trading Strategies  |  Python, Deep Learning, Algorithmic Trading, Pandas, Forecasting

  2022

  • With capabilities to backward or forward test implemented strategies.

  • Implementing different cryptocurrencies trading strategies based on different indicators.

  • Implementing different trading strategies using deep learning methods.

• TradeGate  |  Python, Algorithmic Trading, Pandas

  2022

  • A python library for algorithmic trading and connecting to different exchanges.

  • Unifying different exchanges' APIs.

  • Implemented using different design patterns.

• Automatic Real-Time Refereeing for Bocce Ball Game  |  Python, Computer Vision, Object Detection, Pytorch, OpenCV

  2021

  • Using four wide angle cameras above the center line of the game court.

  • Detecting balls on the ground (four colored balls with a single white ball) using YOLOv5.

  • Finding exact distance between colored balls and the white ball using different geometry and image processing techniques.

  • Implemented mainly using OpenCV functions.

• Finding Sentence Similarities  |  Python, NLP, Django

  2021

  • For farsi sentences, using ParsBERT.

  • Implemented for a knowledge management system.

  • Converting sentences to vectors and finding similar sentences using cosine similarities.

  • Provided REST API endpoints using the Django framework.

• An Automatic License Plate Recognition for Iranian Plates  |  Python, Pytorch, Object Detection, Django

  2020

  • Using YOLOv3 and updated to YOLOv5.

  • Data acquisition was made manually, and different image augmentation techniques were used on the dataset.

  • Image augmentation is done using the "Albumentations" package and editing the labels correctly.

  • Implemented as a two-step detector: First locating the plate and then recognizing plate characters inside the plate area.

  • Provided REST and SOAP API endpoints using the Django framework.

• Training CGAN on Fashion-MNIST dataset  |  Python, Keras, Deep Learning

  2020

  • Training the GAN network and generating new data.

  • Analyzing effects of different loss functions and optimizers in training.

  • Generating different results using different latent space points and decoder networks.

• Training Variational Auto Encoder and Conditional Variational Auto Encoder on MNIST dataset  |  Python, Keras, Deep Learning

  2020

  • Training the encoder and the decoder parts.

  • Visualizing the data in latent space with a scatter plot.

  • Generating different results using different latent space points and decoder networks.

• Text Generation Using RNN Network  |  Python, Keras, NLP

  2020

  • Using Shakespear dataset.

  • LSTM layers for data generation and memory.

• Video Frame Prediction Using CNN and RNN  |  Python, Keras, Image Processing

  2020

  • Predicting future frames of pendulum movement video.

  • 1, 5, 10, 20, 50 frames look-ahead window.

  • Using convolutional layers to implement an encoder and a decoder to reduce the frame's dimensions.

  • Using LSTM layers as the middle part of the network between encoder and decoder to predict future frames.

• Solving Frozen Lake Problem Using Q-Learning and Deep Q-Learning  |  Python, Keras, Reinforcement Learning

  2020

  • Using "FrozenLake v0" environment from gym.

  • Implemented Q-learning method.

  • Implementing Deep Q-Learning with experience replay using keras.

• YOLO Object Detection on Chess Pieces Dataset  |  {Python, Keras, YOLO, Computer Vision, Object Detection

  2020

  • Using YOLO-v5.

• Transfer-Learning with EfficientNet  |  Python, Keras, Deep Learning

  2020

  • A review on EfficientNet novelties and network structures.

  • Running inference using EfficientNet-B0 using Keras.

  • Transfer learning using EfficientNet-B0 top part and custom dense layers to classify two classes from the CIFAR-10 dataset

• Solving Simple Problems using Multi-Layer Perceptron  |  Python, Keras, Neural Networks

  2020

  • Predicting average daily rates of "Hotel Booking Demand" dataset as a regression problem.

  • Predicting labels of CIFAR-10 dataset as a classification problem.

  • Implemented using Keras library.

  • Analyzing effects of using different activation functions, loss functions, optimizers, and evaluation metrics.

  • Providing and implementing a solution for unbalanced data.

• Software testing course projects  |  Java, Software Testing, JUnit

  2020

  • Different testing methods implemented on "Spring PetClinic Sample Application".

  • Testing methods including JUnit, test doubles, code coverage methods, input‌ space partitioning, API testing, mutation testing, UI testing, and BDD.

• Ball on Plate Balancing System  |  Python, OpenCV, Computer Vision, Socket Programming, PID Controller

  2020

  • Using a rectangle plate with four arms and a camera above the plate.

  • Fetching ball position using camera and object detection based on the ball's color.

  • Balancing the ball in the center of the plate using a PID controller and four-step motors.

• Jalas Website  |  Java, Java Spring Framework, SQL, NoSQL, REST API

  2019

  • A website to set meeting dates.

  • Creating users and meetings, adding users to meetings, editing meetings, and notifying meeting invitees through email.

  • Java Spring Framework for the back-end and Reach.js for the front-end.

• Jaboonja  |  Java, Java Spring Framework, HTML, CSS

  2019

  • Implementation of a web application for finding projects to do if you are qualified for them.

  • Implemented in Java(Spring framework) for backend and JavaScript(React.js) for frontend.

• "UCI Heart Disease" data set classification  |  Python, Artificial Intelligence, Machine Learning, Data Science

  2019

  • Using bagging decision tree and random forest methods.

• Saadi and Hafez poem classification  |  Python, Artificial Intelligence, Machine Learning, NLP, Data Science

  2019

  • Pre-processing and using the bag-of-words method on the Saadi and Hafez poems dataset.

  • Classification using naive Bayes method.

• Simple Intelligent Game Players  |  Python, Artificial Intelligence

  2018

  • Checkers player using Mini-Max algorithm with alpha-beta pruning method.

  • PacMan player using different searching algorithms (A*, IDS, BFS) to design an intelligent PacMan in a modified game version.

• Operating Systems Course Projects  |  C, C++, Operating Systems, Socket Programming, Threading, Distributed Systems

  2018

  • Synchronization of MLP Neural Networks Using Semaphores: Using semaphore locks to assure synchronization and output integrity of distinct concurrent processes in each layer of an MLP neural network.

  • Multi-Thread Store Search: Using load balancing and multi-threading through a pipeline and asynchronous forking techniques.

  • Online Battleship Game: Using the C programming language, createing a Battleship game in which the server connects two clients and the game runs in a peer-to-peer manner.

• Image compression and decompression  |  Verilog, C, Computer-Aided Design

  2018

  • With color space conversion (RGB to YUV), downsampling, signal transform, quantization, and lossless coding steps for compression.

• Advanced Programming Course Projects  |  C++, RSDL, Qt

  2016 - 2017

  • Taxiare: A taxi ordering application: Client-Server architecture, Qt application with user administration, pricing computation (using graph search methods), and graphical interface.

  • "Plants Vs. Zombies" Game: Simple graphical implementation, employing the RSDL library for graphical rendering and simple C++ to handle various moving elements.

  • Image Steganography: Using a cipher key to perform both the encryption and decryption of a text message contained within a picture.

• Simple project version control  |  C

  2016

  • With the capabilities of creating users and projects, adding users to projects, versioning projects, and keeping track of the versions.

Skills

Interests