Vision Transformers & Pose Estimation
Trained a transformer-based neural network for pose estimation tasks. Press blue link above for PDF report.
Portfolio
Literature Review: Nonconvex Optimization for Low-rank Matrix Factorization
Wrote a literature review paper on solving matrix factorization problems that are useful for things like matrix sensing, phase retrieval, and PCA-like problems. Press blue link above for PDF report.
Metropolis-Hastings: A Markov Chain Monte Carlo Method
Implemented the McMC algorithm to sample posterior distributions in a linear regression type problem. Press blue link above for PDF report.
ML Reproducibility Challange 2022: Zeroth-Order Optimization Methods (Gradient Free)
Reproducibility is an important task in modern research, for this project, we implemented a few optimization methods to confirm some of the results in the original paper. Press blue link above for PDF report.
Convolutional Neural Networks: Object Detection
Implemented a CNN for object detection tasks with handcrafted layers, FCOS, and more. Press blue link above for more detail.
Generative Adverserial Networks (GANs) & Neural Transfer
GAN-based neural network implementation for style transfer generation. Press blue link above for more details.
Autoencoders & CLIP models
Implemented Autoencoder model for image compression and representational learning. Moreover, the CLIP model is explored for query tasks as shown below. Press blue link above for more details.
Diffusion Models
Diffusion-based neural networks have gained traction in recent years for tasks such as generative AI. In this project, I implement a diffusion-based network for image generation using CIFAR10 dataset. Press blue link above for more details.
Laplacian Pyramids: Image Blending
Laplacian pyramids are useful to capture information of images at various scales. In this project, I use this to blend images together as seen below. Press blue link above for more details.
Panorama Stitching with ORB/SIFT Features
Panorama stitching involves joining images taken at different angles. This is nontrivial since it’s challenging to line them up. By using ORB features (key landmarks), one can use this to find the affine mapping related to the two images as shown below. Press blue link above for more details.
VGG Network: Scene Recognition
In this project, I perform classification (scene recognition) with a VGG-type neural network using the miniplaces dataset to recognize the scene of images. Press blue link above for more details.