Detecting Jute Plant Disease Using Machine Learning

  • Multi-class Support Vector Machine (SVM), Feature Extraction, Hue-based
    Segmentation, Morphological Analysis, Blob Detection, Color Co-occurrence
    Methodology
    paper     code    demo1   demo2

 

Weld Quality Analysis using Deep Learning from Ultrasonic B-scan

  • Computer Vision, Object Detection, Tensorflow, Transfer Learning, Single Shot
    Multibox Detector, Faster R-CNN, GPU Programming, Signal Processing, DB3
    Database, SQLite

    paper     code

 


I Taught these AI to Dance using PoseNet and Tensorflow.js

  • PoseNet, Multi-person Pose Estimation,
    Keypoints Extraction, Tensorflow.js, Javascript,
    Canvas

    code

 

Authenticity Analysis of Opinion and Claim in Online – 3rd Winner, Thales Student Innovation Championship 2018

  • Opinion Mining, Sentiment Analysis, CNN, NLP, End-to-end learning, Android
    Application, t-SNE, Word Embedding, Word2Vec

    project    prototype   news

 

DonateMe: A Fully Functional Donation Website

  • PHP, HTML, CSS, Javascript, MySQL, Xampp, PHPMyAdmin, Bootstrap,
    Web Hosting, Server-side

    code

 

Visual Relationship Track using Deep Learning for Vision and NLP

  • Google AI Open Images, Kaggle Challenge, CNN, LSTM, Recurrent Neural
    Network (RNN), DenseCap model, Show and Tell model

    paper

 

Finding Relevant Biomarkers for Prostate Cancer using Feature Selection and Dimensionality Reduction

    • Gene Expression, PCA, LDA, Visualization, Regression, VarianceThreshold,
      Pearson Correlation, Random Forest Classifier, Support Vector Machine (SVM),
      K-fold Cross Validation

      paper     code

 

Machine learning on Breast Cancer Gene Expression Data

    • Naïve Bayes, Random Forest Classifier, Support Vector Machine (SVM), Random
      Forest Regressor, Support Vector Regressor (SVR), Kernel Ridge Regressor
      (KRR), mRMR, Regularized Linear Model, Recursive Feature Elimination (RFE),
      K-fold Cross Validation, ROC curve, L1-L2 Regularization, ElasticNet

      paper     code

 

Unsupervised Feature Learning using Different Autoencoders

    • Autoencoder, Denoising Autoencoder, 2D, 3D, Temporal and Multi-Modal
      Unsupervised Feature Learning, Spatial Feature Learning, Deep Convolutional
      Autoencoder, Stacked Sparsed Autoencoder, Recursive Autoencoder

      paper