Domain-Aware Quantum Circuit for QML

Designing parameterized quantum circuits (PQCs) that are expressive, trainable, and robust to hardware noise is a central challenge for quantum machine learning (QML) on noisy intermediate-scale quantum (NISQ) devices. We present a Domain-Aware Quantum Circuit (DAQC) that leverages image priors to guide locality-preserving encoding and entanglement via non-overlapping DCT-style zigzag windows. The design employs interleaved encode-entangle-train cycles, where entanglement is applied among qubits hosting neighboring pixels, aligned to device connectivity. This staged, locality-preserving information flow expands the effective receptive field without deep global mixing, enabling efficient use of limited depth and qubits. The design concentrates representational capacity on short-range correlations, reduces long-range two-qubit operations, and encourages stable optimization, thereby mitigating depth-induced and globally entangled barren-plateau effects. We evaluate DAQC on MNIST, FashionMNIST, and PneumoniaMNIST datasets. On quantum hardware, DAQC achieves performance competitive with strong classical baselines (e.g., ResNet-18/50, DenseNet-121, EfficientNet-B0) and substantially outperforming Quantum Circuit Search (QCS) baselines. To the best of our knowledge, DAQC, which uses a quantum feature extractor with only a linear classical readout (no deep classical backbone), currently achieves the best reported performance on real quantum hardware for QML-based image classification tasks. Code and pretrained models are available at: https://github.com/gurinder-hub/DAQC.

💡 Research Summary

The paper introduces a novel parameterized quantum circuit (PQC) architecture called the Domain‑Aware Quantum Circuit (DAQC) designed specifically for image classification on noisy intermediate‑scale quantum (NISQ) devices. Existing PQCs typically employ data‑agnostic entanglement patterns and deep circuits, which quickly become impractical on current hardware due to limited qubit connectivity, SWAP overhead, noise accumulation, and barren‑plateau phenomena. DAQC addresses these issues by embedding strong image‑domain priors directly into the circuit construction.

First, input images are down‑sampled via adaptive average pooling to a compact N × M grid that matches the available qubit budget. The grid is partitioned into non‑overlapping p × q patches. Within each patch a discrete‑cosine‑transform (DCT)‑style zig‑zag scan orders the pixel intensities so that spatially adjacent pixels become consecutive entries in a one‑dimensional vector. After a linear scaling that maps pixel values to rotation angles in the interval