Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Quantum-enhanced satellite image classification

Qi Zhang, Anton Simen, Carlos Flores-Garrigós, Gabriel Alvarado Barrios, Paolo A. Erdman, Enrique Solano, Aaron C. Kemp, Vincent Beltrani, Vedangi Pathak, Hamed Mohammadbagherpoor

Abstract

We demonstrate the application of a quantum feature extraction method to enhance multi-class image classification for space applications. By harnessing the dynamics of many-body spin Hamiltonians, the method generates expressive quantum features that, when combined with classical processing, lead to quantum-enhanced classification accuracy. Using a strong and well-established ResNet50 baseline, we achieved a maximum classical accuracy of 83%, which can be improved to 84% with a transfer learning approach. In contrast, applying our quantum-classical method the performance is increased to 87% accuracy, demonstrating a clear and reproducible improvement over robust classical approaches. Implemented on several of IBM's quantum processors, our hybrid quantum-classical approach delivers consistent gains of 2-3% in absolute accuracy. These results highlight the practical potential of current and near-term quantum processors in high-stakes, data-driven domains such as satellite imaging and remote sensing, while suggesting broader applicability in real-world machine learning tasks.

Quantum-enhanced satellite image classification

Abstract

We demonstrate the application of a quantum feature extraction method to enhance multi-class image classification for space applications. By harnessing the dynamics of many-body spin Hamiltonians, the method generates expressive quantum features that, when combined with classical processing, lead to quantum-enhanced classification accuracy. Using a strong and well-established ResNet50 baseline, we achieved a maximum classical accuracy of 83%, which can be improved to 84% with a transfer learning approach. In contrast, applying our quantum-classical method the performance is increased to 87% accuracy, demonstrating a clear and reproducible improvement over robust classical approaches. Implemented on several of IBM's quantum processors, our hybrid quantum-classical approach delivers consistent gains of 2-3% in absolute accuracy. These results highlight the practical potential of current and near-term quantum processors in high-stakes, data-driven domains such as satellite imaging and remote sensing, while suggesting broader applicability in real-world machine learning tasks.
Paper Structure (8 sections, 2 equations, 3 figures, 1 table)

This paper contains 8 sections, 2 equations, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Methodology
  3. Image feature extraction
  4. Digitized quantum feature extraction
  5. Classification
  6. Results
  7. Image dataset preparation
  8. Discussion and conclusion

Figures (3)

  • Figure 1: Overview of the quantum-enhanced classification pipeline. Multisensor remote-sensing imagery is first mapped to an n-dimensional classical feature vector (in this case, n = 15, 120 and 156). These features parametrize an n-qubit Hamiltonian, which is processed through Digitized Quantum Feature extraction (DQFE) using a counterdiabatic evolution protocol. Local measurements yield one-body expectation values and two-body correlation functions, which are then used as input to a classical classifier such as gradient boosting or random forests.
  • Figure 2: Overview of the reduced TreeSatAI dataset. (a) Representative aerial patches from the five selected classes used in the hardware-compatible benchmark. (b) Example of the three sensing modalities available for each ground location: aerial imagery, Sentinel–2 multispectral measurements, and Sentinel–1 radar data. These heterogeneous inputs are reduced to $n$ classical features before quantum processing.
  • Figure 3: Two-dimensional PCA projections and their mean Fisher information, $\bar{F}$, of the (a) classical and (c) quantum feature representations and the respective classification performances on classical and quantum datasets for the TreeSatAI reduced benchmark.