A Review on Forensic Science and Criminal Investigation Through a Deep Learning Framework

Literature Survey

A comprehensive description and development of a novel method for object detection in images using Deep Neural Networks were provided by Christian et al. In the research, a machine learning model was developed to successfully classify images and localize various object positions detected in the image. The research comprehensively explains how deep neural networks outperformed other classification techniques and it presented neural networks as a more powerful and robust algorithm suitable for classification problems because of its deep architecture. The validity of the approach used in the model developed is analyzed by using it on VOC as the test dataset. The experiment conducted with the model on this dataset utilizes boundary boxes to detect a significant object in these test images and its conclusive accuracy is compared with three related approaches which include: sliding window version of a DNN classifier (Du et. al., 2018), a 3-layer compositional model by (Jin et al., 2010) and the DPM by (Felzenszwalb et al., 2010) and (Girshick et al., 2013) to evaluate the achieved results of the model.