I am looking for bright PhD students. Current openings (x2) are listed below (currently funding is not available). Self-funded students with similar research ideas are welcome to discuss futher (Email: deepayan.bhowmik@shu.ac.uk).

PhD position I: Biologically inspired low power and adaptive computer vision for scene understanding

The high performance demands of signal and image processing (SIP) systems require dedicated hardware system using dedicated system-on-chip solutions, applications processor, graphical processing unit (GPU) or field programmable gate arrays (FPGAs). Deployment scenarios include a range of applications such as mobile robotics, autonomous cars, mobile and wearable devices or public space surveillance (airport / railway station). Modern SIP systems which play a significant role in such interaction process require efficient tools and computing architectures that have ultra-fast processing capabilities operating at extremely low power. Current such systems rely on traditional CPUs and GPUs which were built for general purpose processing (slower response time) and resource hungry (higher power consumption); and hence prone to fail in environments with limited power, bandwidth and computing resources. The aim of this PhD is to conduct research on biologically inspired adaptive computer vision system that is low power, computationally efficient targeting FPGA based heterogeneous computing platforms for scene mapping (visual SLAM).

PhD position II: Visual Attention based techniques in detecting image / video forgery

Dramatic expansion of digital technologies within the last decade resulted in surge in consumption of media content. However it also attracts parallel piracy industry infringing copyrights that costs the digital economy hugely e.g., the cost of piracy in the UK is £500m a year to the audio-visual industries. Further tempering, i.e., adding and removing objects from the scene (digital forgeries) and intentional editing (e.g. Photoshopping for advertisements) pose threat to media authenticity. This has major social impact, e.g., 1) often social unrests (including riots) were observed due to sharing of false / doctored media contents through social networking websites or 2) eating disorders (anorexia) have been reported specially among the teenagers passionately following celebrities with Photoshopped advertisements promoting artificial perfect body shapes etc. The aim of this PhD proposal is to conduct research in watermarking domain and media forensics to address such issues. This includes watermarking based methods for copyright protection using visual attention based approaches and physics based methods (e.g. Lens aberration or directional feature of light sources in images or videos) for forgery detection. Unique media forgery detection techniques will be researched using visual attention and Physics based approaches.