Quantum Computing for Computer Vision: Applications, Challenges, and Research Tracks

In the last few years, computer vision has achieved significant breakthroughs, largely due to the advances in deep learning models. However, despite these remarkable achievements, deep learning models for computer vision are already showing their limits. Quantum computing's capacity for parallelism and complex data processing presents a novel approach to tackling the computational demands of computer vision tasks. In this context, quantum computing emerges as a potential solution to the challenges in computer vision. Quantum principles have the potential to enhance computational efficiency and accuracy, opening doors to new horizons in solving complex computer vision problems. In this paper, we investigate the use of quantum computing for computer vision. First, we analyze quantum architectures and the evolution of quantum computing specifically for computer science. This analysis offers a foundational understanding of quantum computing and quantum techniques. Second, we study applied quantum computing research for computer vision through an extensive literature review. Simultaneously, we present an in-depth analysis of the existing limitations and challenges posed by quantum hardware and algorithms in computer vision applications as well as outline the potential research tracks for applied quantum computing in computer vision.