Anand P V's Projects
ADAS Car - with Collision Avoidance System (CAS) - on Indian Roads using LIDAR-Camera Low-Level Sensor Fusion. DIY Gadget built with Raspberry Pi, RP LIDAR A1, Pi Cam V2, LED SHIM, NCS 2 and accessories like speaker, power bank etc
Math Techniques viz. ARIMA, Frequency Decomposition, Fourier Filtering, Linear Regression & Bi-directional LSTMs on Feature Engineered Stock Market Data.
Machine Automated Reasoning using Grakn KGLib, Knowledge Graph Convolutional Networks (KGCNs) & Machine Comprehension using BERT-BIDAF-Ensemble
Design, Provision and Monitor AWS Infrastructure at Scale. Cloud Architect Nano-degree
A chatbot implemented using RNN and GloVe embeddings whch answers your query crazily
Semantic clustering of text using K-means on custom-built contextual Word Co-occurence matrix, decomposed using SVD. Truncated length of word vectors are estimated by plotting explained variance of decomposed matrix.
To classify images containing handwritten digits using multiple custom-built CNN architectures, which may or may not are inspired from standard Convnet architectures such as LeNet, AlexNet, VGGNet, ResNet etc. A comparison between performance of different architectures are done.
Computer Vision for Skew Correction, Text Inversion, Rotation Classification, Homography & Object Search with Applied Math
Linear Model to predict the crime rate of North Carolina. Detailed EDA done prior to building model
To analyze the performance of DT on BoW, tf-idf, W2V & tf-idf weighted W2V featurizations of Amazon Reviews data. Tree depth is taken as the hyperparameter tuned using GridSearchCV. The similarity in shape of CV Error vs Depth plots across all featurizations is noticeable.
Visual data analysis on Haberman Dataset
Face Recognition-based Attendance system deployable on Raspberry Pi with Movidius NCS2. RPLIDAR A1 is integrated for depth sensing to enable Door Access Control simulated with Pimoroni Blinkt!
Gradient Boosted Decision Tree & Random Forest Classifier on high dimensional data
Gesture Triggered Alarm for Women Security based on Computer Vision, Mathematical Concavity Estimation, MQTT, OpenVINO and Pimoroni Blinkt! on RPi or Jetson Nano
Performance comparison of self-coded Stochastic Gradient Descent vs Sklearn OLS vs Batch Gradient Descent
Traditional ASR (Signal & Cepstral Analysis, DTW, HMM) & DNNs (Custom Models + DeepSpeech) on Indian Accent Speech
Indoor Navigational Assistance with Custom Object Detection and SLAM on Raspberry Pi and Android Mobile
Hyperparameter Tuning of K-Means using Elbow Method, Eps values based on MinPoints for DBScan and Hierarchical Clustering based on various values of 'k' is done to form different clusters, on which an analysis is done for semantic coherence.
10-fold CV KNN on BoW, TF-IDF, W2V & TF-IDF weighted W2V (Brute force vs kd-tree)
Logistic Regression on Amazon Reviews
Personified Generative Chatbot using RNNs (LSTM) & Attention in TensorFlow
MATLAB Code for Linear & Logistic Regression, SVM, K Means and PCA, Neural Networks Learning, Multiclass Classification, Anomaly Detection and Recommender systems.
5-class Sentimental Analysis using various LSTM Architectures to draw comparisons
Efficient Text Localization Algorithm, Image Inversion Detection of Scanned Documents & Language Identification based on Shape Context and Traditional Computer Vision.
To use different MLP architectures having different number of layers on the famous MNIST database. Tweaking methods like dropouts, batch normalization are applied and cross entropy loss is being tracked for analysis and comparison between architectures.
10-fold CV Naive Bayes Classifier on BoW and TF-IDF vectorized Reviews with Optimal Alpha
Starter code for Course 3 of nd063
Enabling HCI on the Edge: Gesture & Offline Sound Control of kiosks with AI models, hardware optimized with Intel OpenVINO, code optimized with VTune
Multi-Class Object Detection on Mobile Video Stream, using Deep Learning ConvNets, to assist the blind or to signal an incoming threat, without radars.
Parallel implementation hack of inherently sequential algorithms. Random Number Generators - Additive LFG and GFSR - implemented with NVIDIA CUDA using Continuous Subsequence Technique and Leap Frog Technique. Paper presented in the International AI Conference 2022, Tel Aviv, Israel.