Internship - Machine Learning Engineer
About the Role
We are looking for a curious, motivated, and data-driven Computer Vision Intern to join our Machine Learning team. During this internship, your primary mission will be to research, design, and prototype a new Scene Classification Model. You will work with real-world image datasets to teach our systems how to automatically categorize and understand visual environments (e.g., distinguishing between indoor, outdoor, urban, nature, or specific room types).
You won't just be fetching coffee; you will be writing production-level code, training models, and presenting your findings to the engineering team. You will be paired with a senior ML engineer who will provide weekly mentorship and guidance.
What You Will Do
- Data Curation & Preprocessing: Help gather, clean, and annotate image datasets required for training the scene classification model. Apply data augmentation techniques to ensure model robustness.
- Model Development: Experiment with state-of-the-art architectures (such as CNNs, ResNets, or Vision Transformers) to build classification model using PyTorch.
- Evaluation & Optimization: Track model performance using metrics like accuracy, precision, recall, and F1-score. Fine-tune hyperparameters to optimize for both speed and accuracy.
- Documentation: Maintain clear notes on your experiments, methodologies, and codebase to ensure a smooth handover at the end of your internship.
Who You Are (Requirements)
- Available for an internship for at least 3 months.
- Currently pursuing a BS in Computer Science, Data Science, Artificial Intelligence, or a related field.
- Solid programming foundation in Python.
- Familiarity with deep learning frameworks like PyTorch.
- Basic understanding of core Computer Vision concepts and image processing (OpenCV, PIL).
- Strong problem-solving skills and a willingness to independently research new ML papers and techniques.
Bonus Points (Nice to Have)
- Previous projects or portfolio (GitHub/Kaggle) showcasing image classification or object detection work.
- Familiarity with Git and version control.
- Exposure to cloud platforms (AWS, GCP, or Azure) or GPU training environments.