# **Personal Meeting 10/26**

## **Research Purpose**
1. Using student dance movements for human pose segmentation for the input for Human pose recognition
Target: This research will explore and manipulate the COCO (Common Objects in Context) image dataset for human video Segmentation. Using Python, PyCoco, and Tensorflow Keras (**still finding the best**).
Dataset : COCO,MPII,Crowdpose - **increase progressing**
Test Data : Dance Student video

3. Using Human pose estimation as recognition from student dance movements for pose classification
Target: This research will explore and manipulate the Human pose estimation as recognition from student dance movements for pose classification (**still finding the best classification methods**).
- find better procedure the Human pose estimation as recognition from other research before
- Manipulate pose classification output become input for assessment

3. Using The pose classification data for the dance assessment result.
Target: This research will explore and manipulate the pose classification comparation as dance assesment according laban notation concept.

Laban analyses; creating an online database for Laban movement studies
http://www.laban-analyses.org/jeffrey/2006-creating-laban-analysis-database/index.htm
GenLaban: A tool for generating Labanotation from motion capture data
https://journals.sagepub.com/doi/abs/10.1177/0539018417694773?journalCode=ssic

## Purpose Research Model

## **This week Logbook**
Using student dance movements for human pose segmentation for the input for Human pose recognition
1. https://datagen.tech/guides/image-datasets/ms-coco-dataset-using-it-in-your-computer-vision-projects/
1. https://opencv.org/introduction-to-the-coco-dataset/
1. https://www.immersivelimit.com/tutorials/create-coco-annotations-from-scratch
1. https://towardsdatascience.com/master-the-coco-dataset-for-semantic-image-segmentation-part-1-of-2-732712631047
1. https://forums.fast.ai/t/image-segmentation-on-coco-dataset-summary-questions-and-suggestions/36816
1. https://majianglin2003.medium.com/object-detection-datasets-coco-ebadb64a21e4
1. https://www.marktechpost.com/2021/03/19/coco-a-definitive-dataset-for-deep-learning-on-images/
1. https://www.tensorflow.org/datasets/catalog/coco

## **Dataset**

Coco - already downloaded and foldering
-Train17 18GB, Val2017 797MB, 2017 Train/Val annotations 246MB & 1,12GB

MPII - already downloaded and foldering
-MPII_human_pose_v1 11.8GB, MPII_human_pose_v1_u12_2 12MB

Crowdpose - already downloaded
-Train + Validation + Test Images 2.2GB, Annotations 110MB.