Amir Ziaee's repositories
Counting-Trees-using-Satellite-Images
This study investigates the aspect of localizing and counting trees using satellite images to create an inventory of incoming and outgoing trees for an annual tree inspections.
Programm-An-Autonomous-Vehicle-in-ROS
In the project I am going to programm a car to drive itself around the test track using components of the perception, planning and control subsystems which will be implemented in the Robot Operation System.
A-new-efficient-lightweight-semantic-segmentation-Network
The E-UNet is a powerful architecture for real-time semantic segmentation, especially if one wants to run it in low-resource embedded systems.
NLP-and-Pipelines
This Repo give you an overview of how to design an end-to-end NLP pipeline.
Recommendation-Systems
In this Repository, I will be working with the MovieTweetings data to create three type of recommendations (Knowledge Based Recommendations, Collaborative Filtering Recommendations and Content Based Recommendations) to recommend movies to users.
Analyse-Disaster-Data
In this Repos I am going to analyze disaster data from Figure Eight to build a natural language classification model for an API that classifies disaster messages.
Generative-Adversarial-Network--GAN-
In this Repo I will explain and code the mathematical principles of the
Prostate-Segmentation
In this repository, I tried to segment prostate images (MRI) using Deep Learning networks.
Face-Detection-and-Verification
In this Repositiory, the goal is to code networks that are developed for Face Verification and Detection.
Machine-Translation
In this notebook, I will build a deep neural network that functions as part of an end-to-end machine translation pipeline. My completed pipeline will accept English text as input and return the Persian translation.
Markov-Localization-and-the-Kidnapped-Vehicle-
Markov Localization or Bayes Filter for Localization is a generalized filter for localization and all other localization approaches are realizations of this approach. By learning how to derive and implement this filter, we develop intuition and methods that will help us solve any vehicle localization task.
Part-of-Speech-Tagging
In this notebook, you'll use the Pomegranate library to build a hidden Markov model for part of speech tagging with a universal tagset.
Sentiment-Analysis
Sentiment Analysis with logistic Regression and Fully Connected layer and LSTM (Tensorflow)
Spam-Classifier-with-Naive-Bayes
In this this repository I will be using the Naive Bayes algorithm to create a model that can classify SMS messages as spam or not spam.
The-Usage-of-AI-defects-for-Building-Footprint-Segmentation
To investigate the effect of pre-processing algorithms in the field of Building Footprint Segmentation, we are going to design a pre-processing algorithm that manipulates the training data to feed them into a Mask R-CNN network that can perform instance segmentation or fake semantic segmentation based on the execution or non-execution of the pre-processing algorithm
Explainability-in-Machine-Learning-Models-for-Credit-approval
Defaulting on credit can significantly impact your ability to borrow money in the future. In this study, we found that a prior default decreases the chance of getting credit by 70 percent. This explanation is based on the analysis of three machine learning models (A, B, and C) which are train on the Credit Approval Dataset.
Object-Detection--MobileNet-and-MobileNetSSD-
Learn about MobileNets and separable depthwise convolutions. and the SSD (Single Shot Detection) architecture used for object detection.
cocalc-docker
Docker setup for running CoCalc as downloadable software on your own computer
Deciphering-Encrypted-Text-with-character-level-RNNs
In this repository I build a recurrent neural network and train it to decipher strings encrypted with a certain cipher.
Find-prime-numbers-by-various-threads
A multi-threaded c program to find prime numbers between two ranges and the time it takes the program to run.
keras-unet-collection
The Tensorflow, Keras implementation of U-net, V-net, U-net++, UNET 3+, Attention U-net, R2U-net, ResUnet-a, U^2-Net, TransUNET, and Swin-UNET with optional ImageNet-trained backbones.
knodle
A PyTorch-based open-source framework that provides methods for improving the weakly annotated data and allows researchers to efficiently develop and compare their own methods.
Swin-Transformer
This is an official implementation for "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows".
Topic-Modeling
In this repository, I'll use the gensim library to build LDA (Latent Dirichlet Allocation) to classify text in a document to a particular topic.