Human Activity Recognition (HAR) plays an important role in many real world applications. Massive augmentation in HAR related applications has been observed over past decade. Currently, various techniques, such as wearable devices and mobile devices have been proposed for sensor-based “HAR” in daily health monitoring, rehabilitative training and disease prevention. However, non-visual sensors in general and wearable sensors in specific have several limitations: acceptability and willingness to use wearable sensors; battery life; ease of use; size and effectiveness of the sensors. Therefore, adopting vision-based human activity recognition approach is more viable option since its diversity would enable the application to be deployed in wide range of domains. Vision-based Human Activity Recognition is the procedure of classifying image sequences with labeling of actions. The classification is tricky in itself because it involves recording performance on the basis of discrepancy in motion performance.
The most popular technique of vision based activity recognition, Deep Learning, however, requires huge domain-specific datasets for training which, is time consuming and expensive. To address this problem we propose a Transfer Learning technique by adopting vision-based approach to “HAR” augmented by already trained Deep Learning models.
A new stochastic model is developed by borrowing the concept of “Dirichlet Allocation” from Latent Dirichlet Allocation (LDA) for an inference of the posterior distribution of the variables relating the deep learning classifiers predicted labels with the corresponding activities. Results show that an average accuracy of 95.43% is achieved during training the model as compared to 74.88 and 61.4% of Decision Tree and SVM respectively. In addition to this, proposed methodology unambiguously addresses the challenges and facilitates an interpret-ability of its functionality as compared to SVM and Neural Networks.
However, testing accuracy suffers due to non-reproducibility of labels by the deep learning models. For detailed insight on “Combining off-the-shelf Image Classifiers with Transfer Learning for Activity Recognition”, by Amit Kumar under guidance of Dr.Mohit Kumar, follow the hyperlink https://dl.acm.org/citation.cfm?id=3266219