Job Description

Duties and responsibilities: 




  





 ● Model Training & Fine-Tuning:
 



 ○ Train and fine-tune machine learning models, including YOLO, ViT, and other advanced deep learning models, specifically for image anomaly and forgery detection tasks.
 



 ○ Experiment with different configurations and hyperparameters to improve the performance of models for detecting image forgeries and anomalies.
 



 ○ Explore and implement techniques for detecting subtle forgery signs, such as image splicing, pixel-level manipulation, and other types of image tampering.
 



  





 ● Data Preprocessing:
 



 ○ Process and clean raw image data for model training, including image augmentation and preprocessing techniques (using OpenCV, PIL, etc.).
 



 ○ Develop and implement feature engineering strategies for improving model input data, ensuring models are trained with the best available features.
 



  





 ● Model Evaluation & Optimization:
 



 ○ Evaluate the performance of machine learning models using appropriate metrics (e.g., precision, recall, F1 score, ROC-AUC) to ensure that models accurately detect anomalies and forgeries.
 



 ○ Continuously fine-tune model performance and optimise for real-time inference and scalability.
 



  





 ● ETL:
 



 ○ Write optimised SQL queries to extract, manipulate, and preprocess large image datasets for training purposes.
 



 ○ Manage and organise large amounts of structured and unstructured data for analysis and model development.
 



  





Skills/ Qualifications: 




  





 ● Qualifications:
 



 ○ Bachelor's or Master's degree in Computer Science, Data Science, Machine Learning, or a related field.
 



 ○ 2+ years of experience in data science with a focus on computer vision, image processing, and machine learning.
 



  





 ● Technical Skills:
 



 ○ Proficiency in Python programming, with experience using libraries such as NumPy, pandas, scikit-learn, OpenCV, and matplotlib.
 



 ○ Strong understanding of machine learning and deep learning fundamentals, including classical models (e.g., decision trees, SVMs) and neural networks.
 



 ○ Hands-on experience with image processing and computer vision tasks, including anomaly detection and image forgery detection.
 



 ○ Familiarity with machine learning frameworks like TensorFlow, PyTorch, or Keras for model building and training.
 



 ○ Familiarity with transfer learning, fine-tuning pre-trained models (e.g., ViT, YOLO), and their applications to anomaly detection and forensics.
 



  





 ● Data extraction:
 



 ○ Experience writing optimised SQL queries to interact with databases and extract relevant data.
 



 ○ Ability to handle and preprocess large datasets for machine learning, including image data.
 



  





 ● Model Evaluation & Performance Metrics:
 



 ○ Strong experience with evaluating models using metrics like accuracy, precision, recall, F1 score, confusion matrix, and ROC-AUC.
 



 ○ Ability to identify and troubleshoot performance issues in machine learning models.