👋 Hi there! I'm Animikh, a Machine Learning Engineer with a passion for Anime and Video Games. I'm currently working as a Computer Vision and Machine Learning Engineer atMoultrie - An EBSCO Company. Here, we're developing next-generation Computer Vision algorithms for Cellular Trail Cameras, aimed at enhancing wildlife monitoring.

I graduated with an MS in AI from Boston University, where I worked under Prof. Eshed Ohn-Bar at theH2X Lab. My research focused on end-to-endAutonomous Driving, specifically on closing the Sim2Real gap and developing offline and online driving evaluation metrics for my thesis.

Previously, I was the Computer Vision Engineer and Lead atWobot.ai, where I spearheaded the development of a robust deep learning tech stack that powers real-time video analytics across hundreds of cameras worldwide.

I ❤️ building things and strongly believe that Multi-Modal Self-Supervised Learning is key to AGI 🤫. My areas of focus include Generative AI, Multi-Modal Learning, and more.

I'm always open to new opportunities and a good chat ☕. Feel free to connect with me onLinkedIn or reach out atanimikhaich@gmail.com.

PinnedLoading

1. VidTuneVidTunePublic

   Forked fromtensorsofthewall/VidTune

   VidTune: Tailored Music For Your Videos

   Python 1

2. No-Code-Classification-ToolkitNo-Code-Classification-ToolkitPublic

   Containerized Tensorflow-based image classification training utility with Streamlit-based interface designed to choose between common architectures and optimizers for quick hyperparameter tuning.

   Python 8 3

3. 3D-Text2LIVE3D-Text2LIVEPublic

   Zero-shot, text-driven appearance manipulation on multiple views of an object to generate 3D renderings.

   Python 3 2

4. Semantic-Segmentation-using-AutoEncodersSemantic-Segmentation-using-AutoEncodersPublic

   Lightweight and Fast Person Segmentation using Autoencoders (Trained Weights Included)

   Jupyter Notebook 20 7

5. ECG-Atrial-Fibrillation-Classification-Using-CNNECG-Atrial-Fibrillation-Classification-Using-CNNPublic

   This is a CNN based model which aims to automatically classify the ECG signals of a normal patient vs. a patient with AF and has been trained to achieve up to 93.33% validation accuracy.

   Jupyter Notebook 48 19

6. Deep-Convolutional-Background-SubtractorDeep-Convolutional-Background-SubtractorPublic

   End-to-end CNN-based Autoencoder that can segment any objects even if it is out of the classes present in the training set.

   Jupyter Notebook 4 1