CVPR Daily - Tuesday

Acknowledging the need for a radical shift in computational paradigm, we have looked into recent breakthroughs in quantum computing . Modern quantum computers (QCs) , capable of leveraging quantum phenomena like superposition, entanglement, and tunneling, are no longer limited to simulations. This opens the door to the exciting field of Quantum Computer Vision (QCV) , which aims to transpose existing computer vision problems into a framework suitable for quantum computation. However, leveraging quantum hardware to tackle complex vision tasks presents unique challenges: − How can we adapt current CVML algorithms to function on QCs? − How can we create hybrid solutions maximizing the strengths of CPUs, GPUs, and QPUs? − How can we devise scalable divide-and-conquer algorithms for optimal QPU utilization? − How can the process of quantum implementation inform CPU implementations ? During our CVPR workshop , Q-CVML, three core themes were discussed through excellent invited talks: (i) Adiabatic quantum computing, presented by Michael Möller from MPI and Victoria Goliber from D-Wave , (ii) Gate- based quantum computing, discussed by Roberto Bondesan from Imperial College and Tat-Jun Chin from University of Adelaide , and (iii) Quantum- inspired computer vision, presented by Anand Rangarajan from University of 17 DAILY CVPR Tuesday QCVML - Quantum CV and ML