I am midway through a Master’s thesis on automatic micro-expression recognition and want to strengthen both the technical experiments and the written document. Everything must be scripted in Python and the final manuscript delivered in LaTeX. Scope • Algorithms: I already hold curated facial micro-expression video sets; your job begins with building and fine-tuning state-of-the-art detection algorithms on these clips. Think CNNs, temporal attention blocks, or any recent lightweight architecture that suits subtle, short-lived motions. • Experiments & Evaluation: Implement robust training and cross-validation pipelines, produce confusion matrices, ROC curves, and statistical significance checks so that the thesis can defend its claims. • Writing in LaTeX: Structure the full document—abstract, related work, methodology, results, discussion and conclusion—directly in a clean LaTeX project (Overleaf-friendly). Embed figures, tables, equations and code listings ready for journal-grade formatting. Expected Deliverables 1. Well-commented Python source (scripts or notebooks) plus requirements.txt. 2. Reproducible model checkpoints and a concise README for rerunning experiments. 3. Complete LaTeX project folder that compiles to PDF without errors. 4. Final PDF draft ready for academic review. Acceptance I will consider the engagement complete once the LaTeX builds without warnings, the metrics in the Results section match the supplied logs, and every figure/code reference resolves correctly. If you enjoy applying computer-vision research to real academic writing, I’m eager to move forward quickly.