Third year university dissertation. Continuous Assessment for ECM3401 - Individual Literature Review and Project. Involves the implementation of a human pose estimation computer vision model to detect two combat athletes, and a machine learning algorithm to identify their grappling position.
Please see specification.pdf
for specification.
The literature review (doc/pdf/lit_review.pdf
) received a final mark of 74/100. The final report and demo (doc/pdf/final_report.pdf
, doc/pdf/final_presentation.pdf
) received a final mark of 76/100. The supervisor report for this project gave a mark of 70/100. This work received a final mark of 75/100.
See src/Pose_Estimation_for_Grappling.ipynb
for source code. Google Colab can be used to load the notebook and run the cells. Please run all cells in the order they appear.
See doc/pdf/
for all documentation files. For LaTeX generated PDFs, the source can be viewed at doc/TeX/
.