Search papers, labs, and topics across Lattice.
The authors introduce M3D-Stereo, a new stereo image dataset designed for training and evaluating image restoration algorithms under multiple realistic degradations. The dataset contains 7904 high-resolution stereo image pairs with pixel-wise aligned ground truths, spanning underwater scatter, haze/fog, underwater low-light, and haze low-light conditions across six degradation levels. Experiments on single-level and mixed-level degradation tasks validate the dataset's utility for benchmarking image restoration and stereo matching methods.
Training robust image restoration models just got easier: M3D-Stereo provides a meticulously controlled, multi-degradation stereo dataset that bridges the gap between synthetic benchmarks and real-world complexity.
Image restoration under adverse conditions, such as underwater, haze or fog, and low-light environments, remains a highly challenging problem due to complex physical degradations and severe information loss. Existing datasets are predominantly limited to a single degradation type or heavily rely on synthetic data without stereo consistency, inherently restricting their applicability in real-world scenarios. To address this, we introduce M3D-Stereo, a stereo dataset with 7904 high-resolution image pairs for image restoration research acquired in multiple media with multiple controlled degradation levels. It encompasses four degradation scenarios: underwater scatter, haze/fog, underwater low-light, and haze low-light. Each scenario forms a subset, and is divided into six levels of progressive degradation, allowing fine-grained evaluations of restoration methods with increasing severity of degradation. Collected via a laboratory setup, the dataset provides aligned stereo image pairs along with their pixel-wise consistent clear ground truths. Two restoration tasks, single-level and mixed-level degradation, were performed to verify its validity. M3D-Stereo establishes a better controlled and more realistic benchmark to evaluate image restoration and stereo matching methods in complex degradation environments. It is made public under LGPLv3 license.
