This thesis studies and evaluates Deep Neural Network models for data demodulation and decoding in a camera-based Visible Light Communication system. Camera communication is an emerging technology that enables communication using light beams, where information is modulated through optical transmissions from light-emitting diodes. This work conducts empirical studies to identify the feasibility and effectiveness of using Deep Learning models to improve signal reception in camera communication. The key contributions of this work include the investigation of transfer learning and customization of existing models to demodulate transmitted signals at the receiver end. The work expounds from a binary quantized system to a 3-bit and 4-bit quantized system. In addition to leveraging Deep Learning methods for demodulating a single VLC transmission, this thesis has developed a pipeline for integration of Deep Learning in a visual multiple-input multiple-output system where transmissions from an LED array are decoded by a camera receiver.