Forensic dentistry focuses on identifying persons based on tooth characteristics or bite mark impressions. This study outlines studies conducted to build an automated method for comparing dental models produced from prospective persons with bite mark impression photos. The best match is then determined by comparing this to a collection of candidate dental models. The findings are compared to those obtained by human forensic odontology specialists. The purpose of this early work was to test the feasibility of utilising Efficient Fast Learning GoogLeNet (EFLG) to identify bite marks using texture data recorded on pink wax and to compare dental cast with pink wax bite marks impression. The experimental findings indicate a fair degree of accuracy.