The paper introduces the Measurement-Adaptive Light-Cone Transform (MALCT), a physics-guided framework that maintains the efficient Fourier-domain inversion of traditional light-cone transform methods while adapting to real-world measurement noise and deviations in non-line-of-sight imaging. MALCT integrates an analytical response with a learned response via measurement-conditioned weighting and predicts a frequency-dependent regularization map within a unified variational formulation. An attention-enhanced spatio-temporal feature extraction module captures long-range temporal dependencies and local spatial correlations, enabling robust reconstruction under low-signal conditions and with reduced acquisition time, as demonstrated on synthetic data and real measurements from various NLOS systems.
arXiv:2609.16017v1 Announce Type: cross Abstract: Time-of-flight non-line-of-sight (NLOS) imaging recovers information from hidden objects by analyzing the time of flight of indirect photons scattere...
STAR (Structure-aware Test-time Adaptation for diffusion-based light field Reconstruction) addresses the ill‑posed problem of reconstructing a light field from limited, noisy focal stack measurements by adapting a pre‑trained diffusion prior to each test scene. The method freezes the diffusion model and trains three lightweight adapters that jointly capture within‑view spatial details, cross‑view angular dependencies, and disparity across views. Experiments show that STAR surpasses state‑of‑the‑art approaches in both two‑ and three‑focal‑sheet settings while achieving faster inference times than methods that update parameters at test time.
Recovering high dynamic range (HDR) radiance from a single standard dynamic range (SDR) image is highly ill-posed. Extreme luminance variation and severe quantization in dark regions make accurate reconstruction challenging, often leading to visual artifacts and color distortions.
arXiv:2609.12682v1 Announce Type: new Abstract: Reconstructing 3D scenes under real-world low-light conditions remains challenging due to severe sensor noise, low signal-to-noise ratios, and degraded...
arXiv:2603.16482v2 Announce Type: replace-cross Abstract: Low-light image enhancement (LLIE) supports optical sensing and scene observation when limited photon capture and sensor noise degrade image...
The paper proposes a method for learning nonlinear waveform fusion across scales to improve integrated sensing and communication (ISAC). By adaptively weighting temporal features and updating spectral components, the authors generate a waveform that maintains fixed total energy while combining image features with a deterministic frequency‑modulated reference. Experiments on 512 test images show that the learned waveforms reduce normalized localization error by 29.0 % and 27.0 % and increase mean peak SNR by 1.40 dB and 0.82 dB compared to baseline approaches.
The paper introduces Gaussian Light Field Splatting (GLFS), a Vision Transformer that embeds continuous light‑field priors derived from 2D Gaussian splatting into its self‑attention mechanism. GLFS models scene illumination as a sum of anisotropic Gaussian basis functions, enabling adaptive spatial gain estimation for low‑light images with complex, non‑uniform lighting. To preserve color fidelity and structural detail, the authors add a color vector angular loss and a luminance edge loss, and report that GLFS outperforms existing methods in illumination correction, structure preservation, and perceptual quality.
Low-light image enhancement is severely ill-posed when the input frame contains missing structure, saturated noise, and weak local contrast. Event cameras provide asynchronous brightness-change observations with high temporal resolution, but prior works often treat voxel channels as an unordered or static feature stack before fusion, rather than explicitly modeling their within-window temporal evolution, weakening the temporal evidence that makes events useful.
arXiv:2606.26738v3 Announce Type: replace Abstract: While recent advancements in generative image editing models have achieved stunning visual fidelity, it remains an open question whether these syst...
arXiv:2608.20602v1 Announce Type: cross Abstract: Many consumer smartphones, stereo cameras, and light field cameras record multiple synchronized viewpoints in a single exposure event. However, novel...
The paper introduces RetinaDiff, a physics‑informed conditional diffusion framework for retinal temporal laser speckle contrast imaging (tLSCI). It first stabilizes speckle sequences via phase‑correlation registration, then uses a motion‑corrected contrast prior to guide a diffusion model that reconstructs high‑quality tLSCI maps from only a few frames. Experiments on an in‑house retinal LSCI system show significant improvements in SSIM, PSNR, and FID compared to conventional five‑frame reconstruction, even in challenging cases.
The paper introduces a two‑stage, projection‑based framework for reconstructing images from only Fourier phase information, using an image‑specific neural prior alongside Fourier‑phase and spatial‑support constraints. In the first stage, the method alternates between enforcing constraints and updating the neural prior, while the second stage refines phase and support with guaranteed convergence. Experiments on 77 microscopy images show that the best variant achieves a pooled PSNR of 31.41 dB, a mean PSNR of 35.75 dB, and a mean SSIM of 0.9531, improving pooled PSNR by 1.51 dB and reducing pooled MSE by 29.3% compared to a constraint‑only baseline.