Here is collection of all the deep learning projects done my me.
An Android App where the user can generate a grid of sticker image, and the user can also save that image for further uses. In the backend GAN model is used to generate new Images and by using rest API those images bring to android app.
Project Source code: Sticker Generator
Notebook: Anim Generator Gan PyTorch
Server Code: Server Code
Multilabel image classification to classify movie poster and find appropriate movie genre with movie poster. For this project tensorflow framwork is used for deep learning model. The model is converted into tf lite to reduce final model size. This is a complete project with server side code and also fronted. To deploy deep learning model, flask is used and deployed into heroku server. By using Flutter framework frontend web app is developed and deployed into firebase.
Project Source code: Multilabel Movie Poster Classification
Notebook: Click here
Web App: Movie Poster Classifier
Server Code: Server code
Flutter web code: Flutter app
The goal of this algorithm is to detect dog breed. If this algorithm detects human, then it will be predicted which dog breed he or she looks like. Or, if detect dog, then it will be predicted dog breed. This project is a part of Udacity Deep Learning Nanodegree.
Project: Dog App
The goal of this project is to predict daily bike rental ridership. In this project, neural network was created from scratch using python and the used dataset contains the number of riders for each hour of each day from January 1, 2011, to December 31, 2012.
Project: Source code
The goal of this project to build a deep neural network that functions as part of an end-to-end automatic speech recognition (ASR) pipeline! The completed pipeline will accept raw audio as input and return a predicted transcription of the spoken language. In this project, different types of architecture are used to find out the best model for this dataset. They are -
- RNN
- RNN + TimeDistributed Dense
- CNN + RNN + TimeDistributed Dense
- Deeper RNN + TimeDistributed Dense
- Bidirectional RNN + TimeDistributed Dense
Notebook Link: Speech Recognition with Neural Networks
A web app where user can draw Bengali digit and the AI model can detect handwritten digit and predict the digit.
Project Source code: DeployBNDegit
Notebook: BN digit with pytorch
In this project, Seinfeld TV scripts was generated using RNNs. You'll be using part of the Seinfeld dataset of scripts from 9 seasons. The Neural Network you'll build will generate a new ,"fake" TV script, based on patterns it recognizes in this training data.
Project Source code: TV Script Generation
The goal of this project is to create a simple web page which a user can use to enter a movie review. The web page will then send the review off to our deployed model (AWS) which will predict the sentiment of the entered review.
Project Source code: Sentiment Analysis Web App
Detect smog in the highway. This project was done as part of Secure and Private AI Scholarship Challenge. This project was selected as the second-best project.
Work On: Model train and testing, Data preprocessing and Datasets (we use our own datasets, every member collect images)
Notebook: Smog Detection
Project Source code: Smog Detection
Pneumonia detection using chest x ray. This project was done as part of Secure and Private AI Scholarship Challenge.
Work on Data preprocessing, model training and testing.
Notebook: chest-x-ray-prediction
Project source code: Chest X Ray Pneumonia
In this project we work with Rice Disease. And compare different CNN Architecture. This project was done as part of Secure and Private AI Scholarship Challenge.
Work on: I use Resnet - 50
Notebook (my version): Notebook
Project Source code: AIMLPros Most
Detect disease using Chest X-ray. This project was done as part of Secure and Private AI Scholarship Challenge.
Work on: Data preprocessing team.
Project Source code: Disease Detection Using chest Xray
Classify fruit images. This project was done as part of Secure and Private AI Scholarship Challenge.
Work on: Model train and testing, Data preprocessing
Project Source code: SG Fruit 360
| Projects Name | Descriptions | Notebook Link |
|---|---|---|
| NLP: EDA + Flair | Top 12 % | Notebook |
| Plant Pathology 2020 - FGVC7 | Top 37 % | Notebook |
| Real or Not? NLP with Disaster Tweets: Disaster NLP: Keras BERT using TFHub | Top 12% | Notebook |
| Titanic: Machine Learning from Disaster | Predict survival number and complete with 0.76 accuracy | Notebook |
| Digit Recognizer | Recognize hand written digit, complete with 0.99717 accuracy | Notebook |
| Projects | Notebook Link | Descriptions |
|---|---|---|
| V2 Plant Seedlings | Notebook | Images of crop and weed seedlings belonging to 12 species. |
| Smog Detection | Notebook | classify image for clear and somg |
| Natural Images classification | Notebook | classify natural image into 8 categories |
| Bengali Number Recognizer | Notebook | Recognizer bengali number |
| Flower Classification | Notebook | Hackathon project, classify 102 types of flower |
| Intel Image Classification | Notebook | Classify 6 types of image |
| Flower Classification | Notebook | classify 5 types of flower image |
| Hot dog or not | Notebook | classify hot dog or not |
| Frute 360 | Notebook | Classify 114 types of fruit |
| Cats and Dogs Classification | Notebook | Classify Cats and dog |
| Flower Classification Udacity final Project | Notebook | Udacity final project |
| Udacity final project with resnet 50 | Notebook | Udacity final project use resnet 50 |
| Malaria detection | Notebook | Import Kaggle dataset to google colab with Kaggle api, classify affected cell of malaria |
| Bird species Classification | Notebook | Classify 14 bard image |
| Project Name | Types | Notebook | Descriptions |
|---|---|---|---|
| Style Transfer | Style transfer | Notebook | Transfer style from one image to another image transfer |
| Style Transfer 2 | Style transfer | Notebook | Transfer style from one image to another image transfer |
| FASHION MINIST | CNN | Notebook | Implement FASHIONMINIST with CNN and FCL |
| Inception Model with Fashion MINIST | CNN | Notebook | Implement Inception Model with FashionMinist |
| CIFAR 10 with tensorboard | FCL | Notebook | Implement linear layer with tensorboard, learn how to use tensorboard |
| Project Name | Notebook | Descriptions |
|---|---|---|
| Celebrity Face Generator | Notebook | Generate fake images form Celebrity datasets |
| Generate Captcha | Notebook | Generate captcha Images |
| Cartoon Character generator | Notebook | Generate cartoon images |
| Project Name | Notebook | Description |
|---|---|---|
| Sentiment Analysis using RNN | Notebook | Sentiment Analysis on Movie Review Dataset |
| Bengali Text Generator | Notebook | Generate text from a book name "Himu" |
| Natural Language Processing with RNN | Notebook | Explore Bag of words, word embedding, RNN, LSTM, and Sentiment Analysis |
| Character Level LSTM in PyTorch | Notebook | Generate English text |
| Notebook Name | Link | Description |
|---|---|---|
| Flair NER with 5M tokens | Notebook | NER using Flair library on 5M word tokens |
| Recommendation | Notebook | Recommendation on Ted talk and Movie using sklean (TF-IDF) |
| Resume Analysis | Notebook | NER on Resume Datasets |
| Spacy NER | Notebook | Spacy: NER, POS, Text preprocessing |
| Spacy Named Extraction | Notebook | Using spacy library, custom NER model is trained |
| Text classification with pytorch | Notebook | Text classification on TREC dataset, using Pytorch |
| Sentiment Analysis with TF | Notebook | Sentiment analysis using Tensorflow |
| Text Classification with TF on TPU | Notebook | Text classification using Tensorflow on Amazon Book review dataset |