Hi, I'm Vedant Agnihotri
B.Tech in Computer Science Engineering | Machine Learning & Computer Vision Enthusiast
I'm a Computer Science Engineering graduate, passionate about developing intelligent systems that solve real-world challenges in healthcare and agriculture. My work centers around Machine Learning, Computer Vision, and AI-driven solutions, aiming to transform industries through innovative technologies. I thrive in interdisciplinary environments where AI intersects with sustainability, precision farming, and medical diagnostics.
About Me
Machine Learning Engineer & AI Researcher specializing in Computer Vision and Agricultural AI
AI-Powered Solutions for Real-World Impact
Hello! I'm Vedant Agnihotri, a Machine Learning Engineer and AI Researcher with a strong focus on Computer Vision and AI applications in agriculture. I'm currently working as a Research Intern at Annam.AI (IIT Ropar), where I'm contributing to innovative solutions using YOLOv8, LangChain, and multimodal learning to address real-world problems in precision farming and sustainability.
I'm passionate about using AI to solve complex challenges in both industry and research. With a background in deep learning, computer vision, and natural language processing, I specialize in developing and deploying models that provide practical insights and real-world impact.
Key Areas of Expertise:
- Computer Vision (CV): Object detection, segmentation, YOLOv8, image classification
- Machine Learning (ML): Supervised learning, CNNs, RNNs, model interpretability
- AI for Agriculture: AI-driven solutions for precision farming, plant disease detection, and leaf counting
- Multimodal Learning: Combining visual and textual data to solve complex tasks
- Deployment & Cloud: Building scalable APIs with FastAPI, deploying models on GCP, using Docker for containerization
In addition to my technical skills, I am driven by a desire to make meaningful contributions to society through AI-powered solutions that push the boundaries of what's possible in fields like agriculture, healthcare, and sustainability.
Feel free to explore my work through my projects below. If you'd like to connect or collaborate, don't hesitate to reach out!
AI Research
Deep learning architectures for medical diagnosis and agricultural applications
Computer Vision
Real-time systems for object detection, segmentation, and gesture recognition
Data Science
Statistical analysis and machine learning model optimization
Software Development
Full-stack development with focus on ML deployment
Experience
Professional journey and research positions
ML Intern
Annam.AI - an AI-CoE of the Ministry of Education, Govt. of India, and DST iHub-AWaDH
Contributing to AI-driven solutions for agriculture and sustainability, with a strong focus on computer vision and deep learning for precision farming
- Developing and optimizing YOLO-based models for leaf counting and crop health monitoring using multimodal data (RGB, thermal, and depth images)
- Leading research on MCLC-NET (Multimodal Continual Learning for Leaf Counting), addressing challenges such as catastrophic forgetting and domain shifts in agricultural datasets
- Implementing robust data pipelines for real-time model deployment and performance evaluation, focusing on scalability and practical application in farming systems
- Collaborating with PhD researchers and cross-functional teams to integrate AI models into real-world agricultural workflows, significantly improving efficiency and sustainability
Summer Research Intern
IIT Ropar
Worked on advanced computer vision research focusing on instance segmentation for agricultural applications
- Developed YOLOv8-based models for crop monitoring
- Implemented COCO-standard annotation pipelines
- Collaborated with PhD researchers on agricultural AI solutions
Machine Learning Intern
Elevate Labs
Developed real-time computer vision applications for accessibility, focusing on ASL gesture recognition
- Architected real-time ASL gesture-to-text conversion system
- Implemented MediaPipe-based hand tracking pipeline
- Achieved cross-platform compatibility with TensorFlow deployment
Technical Skills
Comprehensive overview of my technical expertise
MY SKILLS
Programming Languages
ML/AI Frameworks
Data Science Tools
Web Development
Developer Tools
Areas of Expertise
Featured Projects
Showcasing innovative solutions in ML and computer vision
Respiratory Sound Classification System
Advanced deep learning system for medical diagnosis using spectrogram-based audio analysis
Key Achievements
- Achieved 81.6% accuracy, 76.25% sensitivity, and 92.32% specificity
- Outperformed baseline models by 5.21%
- Surpassed published benchmarks with CNN-LSTM hybrid architecture
- Multi-class classification of four respiratory sound categories
Technologies
Real-Time ASL Gesture Recognition
Computer vision application for American Sign Language gesture-to-text conversion with cross-platform compatibility
Key Achievements
- Real-time hand tracking and gesture recognition
- Custom CNN model with 3 convolutional layers
- Integrated text-to-speech functionality
- Cross-platform compatibility achieved
Technologies
Zucchini Leaf Instance Segmentation
Multi-condition agricultural computer vision system with comprehensive deep learning pipeline for crop monitoring
Key Achievements
- 96.53% mAP@0.5 detection accuracy (13-26% above industry standard)
- 95.71% segmentation quality with 87.94% localization precision
- 50-75% faster training efficiency (<50 epochs vs 100-200 standard)
- Robust performance across varying environmental conditions
- COCO format standardization with comprehensive annotation pipeline
Technologies
Research Publications
Contributing to the advancement of agricultural AI
MCLC-NET: Multimodal Continual Learning for Leaf Counting
A novel approach to solving leaf counting in agriculture using multimodal continual learning. The paper introduces the MMLC dataset for continual learning tasks, tackling challenges such as catastrophic forgetting in real-world farming environments. The model integrates RGB, depth, and thermal imagery to deliver high-accuracy, robust performance. This work provides the foundation for scalable, real-time agricultural monitoring systems.
Respiratory Sound Classification using Deep Learning Approaches
A comprehensive study on using CNN-LSTM architectures with spectrogram analysis for automated respiratory disease diagnosis. Our hybrid model achieved 81.6% accuracy and 92.32% specificity, outperforming baseline models by 5.21% and surpassing published benchmarks.
Get in Touch
Let's discuss your next AI project or research collaboration
Send me a message
I'd love to hear about your project or opportunity
Let's Connect
Download Resume
Get a detailed overview of my experience, skills, and achievements in machine learning and computer vision.