Lab 5 - PoojaShekhar/CS5543-Real-Time-Big-Data-Analytics--Lab-assignments GitHub Wiki

Question Development of Video Classification Application Implement a video application similar to Lab 4 for an interesting use-case related to your project. Use your own Classification Model based on Random Forest / Decision Tree built on Spark MLLib and annotate the frames (for any given video or webcam) with bounding box and text. Report F-Measure, Precision, Recall and Confusion Matrix.

We provided car,home and person as a model.https://github.com/PoojaShekhar/CS5543-Real-Time-Big-Data-Analytics--Lab-assignments/blob/master/Documentation/Lab5/car.txt were used by Decision Tree and RF to detect the objects in the video.

Home-Object Detection

Car-Object Detection

Decision Tree

F-MEASURE: 0.5
RECALL: 0.3333333333333333
PRECISION: 1.0
Accuracy : 0.333

Random Forest

F-MEASURE: 0.5
RECALL: 0.3333333333333333
PRECISION: 1.0
Accuracy : 0.333