Shrawan Kumar

Narayana Darapaneni, Shrawan Kumar, Selvarangan Krishnan et al.

The purpose of this work is, to provide a YOLOv5 deep learning-based social distance monitoring framework using an overhead view perspective. In addition, we have developed a custom defined model YOLOv5 modified CSP (Cross Stage Partial Network) and assessed the performance on COCO and Visdrone dataset with and without transfer learning. Our findings show that the developed model successfully identifies the individual who violates the social distances. The accuracy of 81.7% for the modified bottleneck CSP without transfer learning is observed on COCO dataset after training the model for 300 epochs whereas for the same epochs, the default YOLOv5 model is attaining 80.1% accuracy with transfer learning. This shows an improvement in accuracy by our modified bottleneck CSP model. For the Visdrone dataset, we are able to achieve an accuracy of upto 56.5% for certain classes and especially an accuracy of 40% for people and pedestrians with transfer learning using the default YOLOv5s model for 30 epochs. While the modified bottleneck CSP is able to perform slightly better than the default model with an accuracy score of upto 58.1% for certain classes and an accuracy of ~40.4% for people and pedestrians.

Shrawan Kumar

Although slicing removes code which has no bearing on property checking. However even after that, our study has found that there are predicates in program which have no bearing on property validation, although slicing could not eliminate them. We have cope up with a criteria to identify such predicates and then give a process to leverage them in scale up of property checking.

Shrawan Kumar, Amitabha Sanyal, Uday Khedker

Backward slicing has been used extensively in program understanding, debugging and scaling up of program analysis. For large programs, the size of the conventional backward slice is about 25% of the program size. This may be too large to be useful. Our investigations reveal that in general, the size of a slice is influenced more by computations governing the control flow reaching the slicing criterion than by the computations governing the values relevant to the slicing criterion. We distinguish between the two by defining data slices and control slices both of which are smaller than the conventional slices which can be obtained by combining the two. This is useful because for many applications, the individual data or control slices are sufficient. Our experiments show that for more than 50% of cases, the data slice is smaller than 10% of the program in size. Besides, the time to compute data or control slice is comparable to that for computing the conventional slice.