Quantum Inception Score

Abstract

Motivated by the great success of classical generative models in machine learning, enthusiastic exploration of their quantum version has recently started. To depart on this journey, it is important to develop a relevant metric to evaluate the quality of quantum generative models; in the classical case, one such example is the (classical) inception score (cIS). In this paper, as a natural extension of cIS, we propose the quantum inception score (qIS) for quantum generators. Importantly, qIS relates the quality to the Holevo information of the quantum channel that classifies a given dataset. In this context, we show several properties of qIS. First, qIS is greater than or equal to the corresponding cIS, which is defined through projection measurements on the system output. Second, the difference between qIS and cIS arises from the presence of quantum coherence, as characterized by the resource theory of asymmetry. Third, when a set of entangled generators is prepared, there exists a classifying process leading to the further enhancement of qIS. Fourth, we harness the quantum fluctuation theorem to characterize the physical limitation of qIS. Finally, we apply qIS to assess the quality of the one-dimensional spin chain model as a quantum generative model, with the quantum convolutional neural network as a quantum classifier, for the phase classification problem in the quantum many-body physics.

very close to the published version

References in corpus (57)

• Quantum Machine Learning
• Quantifying Coherence
• An Introduction to Variational Autoencoders
• Quantum metrology
• Barren plateaus in quantum neural network training landscapes
• Quantum Convolutional Neural Networks
• The Role of Relative Entropy in Quantum Information Theory
• Cost Function Dependent Barren Plateaus in Shallow Parametrized Quantum Circuits
• Tensor-Entanglement-Filtering Renormalization Approach and Symmetry Protected Topological Order
• Symmetry protection of topological order in one-dimensional quantum spin systems
• Description of quantum coherence in thermodynamic processes requires constraints beyond free energy
• Superadditivity of communication capacity using entangled inputs
• Quantum generative adversarial learning
• Quantum advantage in learning from experiments
• Generalized Jarzynski Equality under Nonequilibrium Feedback Control
• Quantum generative adversarial networks
• Quantum autoencoders for efficient compression of quantum data
• Quantum coherence, time-translation symmetry and thermodynamics
• The resource theory of quantum reference frames: manipulations and monotones
• Quantum convolutional neural network for classical data classification
• Quantum Boltzmann Machine
• Extending Noether's theorem by quantifying the asymmetry of quantum states
• A Survey on Quantum Channel Capacities
• Hierarchical quantum classifiers
• Experimental Quantum Generative Adversarial Networks for Image Generation
• Quantum speed limits, coherence and asymmetry
• Measuring the quality of a quantum reference frame: the relative entropy of frameness
• Symmetry protected topological phases from decorated domain walls
• Quantum machine learning beyond kernel methods
• The theory of manipulations of pure state asymmetry: basic tools and equivalence classes of states under symmetric operations
• Inequalities for Quantum Entropy: A Review with Conditions for Equality
• Tomography and Generative Data Modeling via Quantum Boltzmann Training
• Quantum Variational Autoencoder
• Machine learning meets quantum physics
• Tradeoff between extractable mechanical work, accessible entanglement, and ability to act as a reference system, under arbitrary superselection rules
• On barren plateaus and cost function locality in variational quantum algorithms
• Entanglement constrained by superselection rules
• Machine learning quantum states in the NISQ era
• Experimental Realization of a Quantum Autoencoder: The Compression of Qutrits via Machine Learning
• Variational Quantum Boltzmann Machines
• The WAY theorem and the quantum resource theory of asymmetry
• Enhancing Generative Models via Quantum Correlations
• Holevo's bound from a general quantum fluctuation theorem
• Machine Learning Topological Phases with a Solid-state Quantum Simulator
• The asymmetry properties of pure quantum states
• Skew informations from an operational view via resource theory of asymmetry
• An Information--Theoretic Equality Implying the Jarzynski Relation
• Quantum Computational Phase Transition in Combinatorial Problems
• Quantum scars and bulk coherence in a symmetry-protected topological phase
• Capacity of Quantum Channels Using Product Measurements
• Classical capacity of the generalized Pauli channels
• Jarzynski-like Equality of Nonequilibrium Information Production Based on Quantum Cross Entropy
• Experimental demonstration of adversarial examples in learning topological phases
• Classical capacity of Gaussian communication under a single noisy channel
• Quantum generative model for sampling many-body spectral functions
• Generative model for learning quantum ensemble via optimal transport loss
• Smooth Metric Adjusted Skew Information Rates

Cited by in corpus (2)

• No-go theorem for environment-assisted invariance in non-unitary dynamics
• Information Processing in Quantum Thermodynamic Systems: an Autonomous Hamiltonian Approach