Tao Yue
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Tao Yue has a total of 66 co-authors. Here are the top co-authors by number of shared publications:
- Xuemei Hu
School of Electronic Science and Engineering Nanjing University Nanjing China
11 shared publications - Feng Yan
Soochow University
3 shared publications - Ting Xu
National Laboratory of Solid-State Microstructures, College of Engineering and Applied Sciences and Collaborative Innovation Center of Advanced Microstructures, Nanjing University 1 , Nanjing 210093
3 shared publications Xuemei Hu
3 shared publicationsXun Cao
3 shared publications- Amit Agrawal
Department of Engineering, University of Cambridge, Cambridge CB3 0FA, United Kingdom
2 shared publications
Latest Publications
- 18 November 2026
- Published by Wiley in Laser & Photonics Reviews
OAM‐Conditioned Speckle Encoding for Single‐Shot Dynamic Imaging
Single‐shot dynamic imaging requires distinguishable, reproducible, and physically implementable temporal encoding states within one camera exposure. However, existing optical encoders often face a trade‐off among encoding diversity, controllability, and stability, limiting practical scalability. In this work, we establish OAM‐conditioned speckle encoding as a physical framework that converts programmable wavefront control into measured intensity‐domain encoding. Programmable orbital‐angular‐momentum (OAM) wavefronts are used as controllable inputs to a fixed scattering medium, which maps programmable OAM states into measured speckle codewords under a fixed, calibrated optical configuration. The resulting measured codebook serves both as temporal modulation basis for encoding a dynamic scene into one snapshot and as the matched physical condition for temporal reconstruction. Numerical simulations and optical experiments demonstrate successful eight‐frame reconstruction, including real single‐exposure measurements of moving targets, with performance characterized under varying compression ratios, measurement perturbations, and cross‐dataset generalization conditions. Codebook mismatch analyses further verify that valid recovery requires the correct correspondence between measured speckle codewords and temporal frames. Overall, OAM‐conditioned scattering establishes a programmable measured‐codebook encoding framework for snapshot dynamic imaging, matched reconstruction, and codebook‐dependent information protection.
- 18 November 2026
- Published by Wiley in Laser & Photonics Reviews
OAM‐Conditioned Speckle Encoding for Single‐Shot Dynamic Imaging
Single‐shot dynamic imaging requires distinguishable, reproducible, and physically implementable temporal encoding states within one camera exposure. However, existing optical encoders often face a trade‐off among encoding diversity, controllability, and stability, limiting practical scalability. In this work, we establish OAM‐conditioned speckle encoding as a physical framework that converts programmable wavefront control into measured intensity‐domain encoding. Programmable orbital‐angular‐momentum (OAM) wavefronts are used as controllable inputs to a fixed scattering medium, which maps programmable OAM states into measured speckle codewords under a fixed, calibrated optical configuration. The resulting measured codebook serves both as temporal modulation basis for encoding a dynamic scene into one snapshot and as the matched physical condition for temporal reconstruction. Numerical simulations and optical experiments demonstrate successful eight‐frame reconstruction, including real single‐exposure measurements of moving targets, with performance characterized under varying compression ratios, measurement perturbations, and cross‐dataset generalization conditions. Codebook mismatch analyses further verify that valid recovery requires the correct correspondence between measured speckle codewords and temporal frames. Overall, OAM‐conditioned scattering establishes a programmable measured‐codebook encoding framework for snapshot dynamic imaging, matched reconstruction, and codebook‐dependent information protection.
- 11 November 2026
- Published by Elsevier in Optics and Lasers in Engineering
- Vol. 209
Physics-coupled neural fields for joint self-calibration in multi-wavelength lensless holography
- 11 November 2026
- Published by Elsevier in Optics and Lasers in Engineering
- Vol. 209
Physics-coupled neural fields for joint self-calibration in multi-wavelength lensless holography
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 48 (11) , 15470-15487
High Frame Rate iToF-Flow-Based Depth Imaging Enabled by Local Linear Transfer
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 48 (11) , 15470-15487
High Frame Rate iToF-Flow-Based Depth Imaging Enabled by Local Linear Transfer
- Open Access
- 26 October 2026
- Published by Wiley in Advanced Functional Materials
- p. e79289
Dynamically Modulated Ultralong Organic Phosphorescent Materials: Achieving Wide‐Range Lifetime Tunability for Multicolor Data Storage
Ultralong organic phosphorescence (UOP) materials have attracted growing research attention in the field of optical information storage, anti‐counterfeiting, and data encryption owing to their unique long‐lived emission characteristics. However, the widespread application of UOP‐based information storage is currently constrained by two main limitations: the dominant binary (on/off) encoding mode with limited information capacity, and the lack of high‐precision, facile data‐writing approaches compatible with scalable fabrication. In this work, we develop a time‐resolved information storage strategy based on widely and precisely tunable UOP lifetimes, which is realized by rationally controlling the polymerization degree of the polymer host. Combining with digital processing (DLP) 3D printing method, the photosensitive materials 3‐TBAOF@HEA:AA and 4‐TBAOF@HEA:AA are designed and developed with tunable UOP lifetimes from 4.6 to 1577.50 ms. Leveraging this tunability of broad lifetime, we demonstrate time‐dependent information storage including multicolor information display and direct photopatterning on flexible 2D thin films. The 3D‐printable UOP system with precisely controllable phosphorescence lifetimes provides a viable material platform for constructing high‐level data security storage devices.
- 26 October 2026
- Published by Wiley in Advanced Functional Materials
- p. e79289
Open AccessDynamically Modulated Ultralong Organic Phosphorescent Materials: Achieving Wide‐Range Lifetime Tunability for Multicolor Data Storage
Ultralong organic phosphorescence (UOP) materials have attracted growing research attention in the field of optical information storage, anti‐counterfeiting, and data encryption owing to their unique long‐lived emission characteristics. However, the widespread application of UOP‐based information storage is currently constrained by two main limitations: the dominant binary (on/off) encoding mode with limited information capacity, and the lack of high‐precision, facile data‐writing approaches compatible with scalable fabrication. In this work, we develop a time‐resolved information storage strategy based on widely and precisely tunable UOP lifetimes, which is realized by rationally controlling the polymerization degree of the polymer host. Combining with digital processing (DLP) 3D printing method, the photosensitive materials 3‐TBAOF@HEA:AA and 4‐TBAOF@HEA:AA are designed and developed with tunable UOP lifetimes from 4.6 to 1577.50 ms. Leveraging this tunability of broad lifetime, we demonstrate time‐dependent information storage including multicolor information display and direct photopatterning on flexible 2D thin films. The 3D‐printable UOP system with precisely controllable phosphorescence lifetimes provides a viable material platform for constructing high‐level data security storage devices.
- 24 September 2026
- Published by Wiley in Laser & Photonics Reviews
- p. e71910
Snapshot Hyperspectro‐Polarimetric Imaging via Disordered Metasurface
High‐dimensional light fields encode rich physical information across spectral, polarimetric, and phase dimensions, far exceeding the intensity‐only data captured by conventional sensors. Accessing these additional domains provides a transformative foundation for multidimensional sensing and intelligent perception, with wide‐ranging applications from biomedical imaging to autonomous systems. However, existing high‐dimensional light‐field imaging systems are often constrained by bulky form factors, limited imaging speed, and limited robustness. This paper presents a novel collaborative framework that integrates a disordered metasurface array with a spatial–spectral–polarization network (SSP‐Net), enabling single‐shot, high‐resolution hyperspectral and full‐Stokes polarimetric imaging. Experimental results demonstrate that the proposed system achieves high‐fidelity high‐dimensional light‐field reconstruction across the visible spectrum (450–650 nm) at a compression ratio of up to 84:1, corresponding to the 21 spectral channels and four full‐Stokes parameters from a single measurement, with a mean peak signal‐to‐noise ratio (PSNR) of 37.11 dB and a mean squared error (MSE) of 0.029 for the degree of polarization (DoP). This work establishes a compact, efficient, and high‐quality paradigm for the acquisition of multidimensional optical information.
- 24 September 2026
- Published by Wiley in Laser & Photonics Reviews
- p. e71910
Snapshot Hyperspectro‐Polarimetric Imaging via Disordered Metasurface
High‐dimensional light fields encode rich physical information across spectral, polarimetric, and phase dimensions, far exceeding the intensity‐only data captured by conventional sensors. Accessing these additional domains provides a transformative foundation for multidimensional sensing and intelligent perception, with wide‐ranging applications from biomedical imaging to autonomous systems. However, existing high‐dimensional light‐field imaging systems are often constrained by bulky form factors, limited imaging speed, and limited robustness. This paper presents a novel collaborative framework that integrates a disordered metasurface array with a spatial–spectral–polarization network (SSP‐Net), enabling single‐shot, high‐resolution hyperspectral and full‐Stokes polarimetric imaging. Experimental results demonstrate that the proposed system achieves high‐fidelity high‐dimensional light‐field reconstruction across the visible spectrum (450–650 nm) at a compression ratio of up to 84:1, corresponding to the 21 spectral channels and four full‐Stokes parameters from a single measurement, with a mean peak signal‐to‐noise ratio (PSNR) of 37.11 dB and a mean squared error (MSE) of 0.029 for the degree of polarization (DoP). This work establishes a compact, efficient, and high‐quality paradigm for the acquisition of multidimensional optical information.
- 17 August 2026
- Published by AIP Publishing in Applied Physics Letters
- Vol. 129 (7)
Super-resolving lensless microscopy via optimized multi-depth fractional Talbot modulation
High-throughput fluorescence imaging is central to fields ranging from immunomics and functional genomics to neuroscience. Lensless microscopy offers an attractive route to meeting such demands through its superior space-bandwidth product (SBP). This framework achieves micrometer-scale resolution across a wide field of view, spanning millimeters to centimeters. However, the resolution of lensless fluorescence microscopy is subject to multifaceted physical and hardware constraints that make it difficult to achieve high-fidelity imaging at the scale required for biological observation. We present spatially encoded high-resolution lensless imaging (Se-HLi), a computational framework that enhances the resolution of compact lensless architectures. Se-HLi uses a movable diffraction grating to generate multi-depth fractional Talbot modulation, thereby encoding high-spatial-frequency information into measurable sensor patterns. A physics-informed spatial modulator aware resolution-improvement transformer then recovers super-resolved details from these measurements. Through differentiable end-to-end optimization of both grating positions and network parameters, Se-HLi improves the system resolution from 5.6 to 3.1 μm (full width at half maximum), while maintaining a minimal hardware footprint. Experimental results show a nearly twofold improvement in resolution, accompanied by a more than threefold increase in SBP. The Se-HLi framework offers a route toward compact, high-throughput platforms for wide-field fluorescence imaging.
- 17 August 2026
- Published by AIP Publishing in Applied Physics Letters
- Vol. 129 (7)
Super-resolving lensless microscopy via optimized multi-depth fractional Talbot modulation
High-throughput fluorescence imaging is central to fields ranging from immunomics and functional genomics to neuroscience. Lensless microscopy offers an attractive route to meeting such demands through its superior space-bandwidth product (SBP). This framework achieves micrometer-scale resolution across a wide field of view, spanning millimeters to centimeters. However, the resolution of lensless fluorescence microscopy is subject to multifaceted physical and hardware constraints that make it difficult to achieve high-fidelity imaging at the scale required for biological observation. We present spatially encoded high-resolution lensless imaging (Se-HLi), a computational framework that enhances the resolution of compact lensless architectures. Se-HLi uses a movable diffraction grating to generate multi-depth fractional Talbot modulation, thereby encoding high-spatial-frequency information into measurable sensor patterns. A physics-informed spatial modulator aware resolution-improvement transformer then recovers super-resolved details from these measurements. Through differentiable end-to-end optimization of both grating positions and network parameters, Se-HLi improves the system resolution from 5.6 to 3.1 μm (full width at half maximum), while maintaining a minimal hardware footprint. Experimental results show a nearly twofold improvement in resolution, accompanied by a more than threefold increase in SBP. The Se-HLi framework offers a route toward compact, high-throughput platforms for wide-field fluorescence imaging.
- 5 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. PP (10518215) , 1
HCDi-Fusion: Hybrid-Conditioned Diffusion Model for Cross-Resolution Multimodal Image Fusion
In multimodal image fusion tasks, introducing diffusion models faces a critical challenge due to the lack of ground truth (GT) for step-wise supervision during the diffusion process. Existing methods typically unfold diffusion models in two ways: either supervising only the final fused result while neglecting intermediate diffusion steps, or adopting mask-based fusion strategies that rely on pre-generated masks to supervise intermediate predicted masks. However, the latter heavily depends on the quality of the pre-generated masks. Moreover, in real-world scenarios, the spatial resolutions of different modalities often exhibit significant discrepancies, which further complicates the generation of reliable pre-generated masks. To address these challenges, we propose HCDi-Fusion for cross-resolution multimodal image fusion. Specifically, the Hybrid-Conditioned Diffusion (HCD) model performs guided denoising by jointly leveraging self-conditioning and external conditioning, which considers inter-step relationships and different resolution discrepancies. Furthermore, the proposed Self-conditioned Step-Wise Fusion (SSWF) model enables progressive fusion in the latent feature space across diffusion steps. To accommodate the inherently unfolded nature of diffusion models in fusion tasks, we further design a Step-Wise Supervision Mechanism to supervise intermediate diffusion steps. Extensive experiments demonstrate that our method achieves state-of-the-art performance across multiple fusion tasks, diverse resolution discrepancies, multiple datasets, and a wide range of evaluation metrics.
- 5 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. PP (10518215) , 1
HCDi-Fusion: Hybrid-Conditioned Diffusion Model for Cross-Resolution Multimodal Image Fusion
In multimodal image fusion tasks, introducing diffusion models faces a critical challenge due to the lack of ground truth (GT) for step-wise supervision during the diffusion process. Existing methods typically unfold diffusion models in two ways: either supervising only the final fused result while neglecting intermediate diffusion steps, or adopting mask-based fusion strategies that rely on pre-generated masks to supervise intermediate predicted masks. However, the latter heavily depends on the quality of the pre-generated masks. Moreover, in real-world scenarios, the spatial resolutions of different modalities often exhibit significant discrepancies, which further complicates the generation of reliable pre-generated masks. To address these challenges, we propose HCDi-Fusion for cross-resolution multimodal image fusion. Specifically, the Hybrid-Conditioned Diffusion (HCD) model performs guided denoising by jointly leveraging self-conditioning and external conditioning, which considers inter-step relationships and different resolution discrepancies. Furthermore, the proposed Self-conditioned Step-Wise Fusion (SSWF) model enables progressive fusion in the latent feature space across diffusion steps. To accommodate the inherently unfolded nature of diffusion models in fusion tasks, we further design a Step-Wise Supervision Mechanism to supervise intermediate diffusion steps. Extensive experiments demonstrate that our method achieves state-of-the-art performance across multiple fusion tasks, diverse resolution discrepancies, multiple datasets, and a wide range of evaluation metrics.
- 29 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. PP (01628828) , 1-16
Learned Optimal Visual Time-of-Flight Imaging with Fisher Information Guidance
Indirect time-of-flight (iToF) imaging provides absolute depth by encoding light transport delays into correlation measurements via active illumination modulation and demodulation. This technology plays a crucial role in object detection and scene understanding and has been widely adopted in applications such as robotics, automotive driving, and augmented reality. However, iToF systems often suffer from low signal-to-noise ratio (SNR), particularly under strong ambient illumination. Signal attenuation and noise-induced phase errors severely degrade depth accuracy, while reliable depth discontinuity and edge reconstruction remain challenging. In this paper, we introduce an optimal visual ToF imaging scheme, optimized through end-to-end learning of iToF coding functions and depth reconstruction, guided by discriminative Fisher information supervision. By integrating vision information from RGB images, we facilitate the convergence of the optimization for iToF coding functions and enhance robustness under noisy inputs. Moreover, we design a dual-branch visual ToF reconstruction network that estimates depth from iToF measurements while exploiting edge guidance derived from visual information to preserve fine geometric details. Extensive experiments on both synthetic and real-world datasets verify the effectiveness of the proposed approach, particularly in challenging low-SNR conditions.
- 29 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. PP (01628828) , 1-16
Learned Optimal Visual Time-of-Flight Imaging with Fisher Information Guidance
Indirect time-of-flight (iToF) imaging provides absolute depth by encoding light transport delays into correlation measurements via active illumination modulation and demodulation. This technology plays a crucial role in object detection and scene understanding and has been widely adopted in applications such as robotics, automotive driving, and augmented reality. However, iToF systems often suffer from low signal-to-noise ratio (SNR), particularly under strong ambient illumination. Signal attenuation and noise-induced phase errors severely degrade depth accuracy, while reliable depth discontinuity and edge reconstruction remain challenging. In this paper, we introduce an optimal visual ToF imaging scheme, optimized through end-to-end learning of iToF coding functions and depth reconstruction, guided by discriminative Fisher information supervision. By integrating vision information from RGB images, we facilitate the convergence of the optimization for iToF coding functions and enhance robustness under noisy inputs. Moreover, we design a dual-branch visual ToF reconstruction network that estimates depth from iToF measurements while exploiting edge guidance derived from visual information to preserve fine geometric details. Extensive experiments on both synthetic and real-world datasets verify the effectiveness of the proposed approach, particularly in challenging low-SNR conditions.
- 6 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36 (7) , 10129-10143
OnRef-VSR: Real-Time Online Reference-Based Video Super-Resolution via Efficient Temporal and Spatial Feature Transfers
Acquiring high-resolution video is essential in applications such as surveillance, remote sensing, and microscopy. Yet pushing spatial resolution often compromises temporal resolution due to readout and bandwidth limitations. Rather than relying solely on hardware improvements, reference-based video superre-solution (Ref-VSR) provides a promising solution to achieve HR video acquisition with off-the-shelf cameras. However, existing Ref-VSR approaches are predominantly offline, relying on future frames and computationally expensive matching, which fundamentally limits their causal real-time applicability. In addition, the inherent trade-off between reconstruction fidelity and inference speed remains a key obstacle to practical use. To overcome these challenges, we present the first online reference-based video super-resolution framework (OnRef-VSR), which enables causal real-time reconstruction through efficient spatial and temporal feature transfers. The framework incorporates three key modules: Flow-Local Linear Transform-based Feature Enhancement (FLFE) module, Hierarchical Heterogeneous Feature Transfer (HHFT) module, and Detail Refinement with Flow-guided Local nonlinear transform (DRFL) module, which in combination realize effective integration of temporal cues and reference information. Extensive experiments demonstrate the superiority of the proposed approach. Specifically, OnRefVSR achieves a 5.2× faster speed with a 0.29 dB gain over the state-of-the-art real-time method, and its offline variant yields a 3.6× speedup and a 0.84 dB PSNR gain on the standard Vid4 benchmark, confirming both the effectiveness and practicality of OnRef-VSR for real-time high-fidelity video reconstruction.
- 6 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 36 (7) , 10129-10143
OnRef-VSR: Real-Time Online Reference-Based Video Super-Resolution via Efficient Temporal and Spatial Feature Transfers
Acquiring high-resolution video is essential in applications such as surveillance, remote sensing, and microscopy. Yet pushing spatial resolution often compromises temporal resolution due to readout and bandwidth limitations. Rather than relying solely on hardware improvements, reference-based video superre-solution (Ref-VSR) provides a promising solution to achieve HR video acquisition with off-the-shelf cameras. However, existing Ref-VSR approaches are predominantly offline, relying on future frames and computationally expensive matching, which fundamentally limits their causal real-time applicability. In addition, the inherent trade-off between reconstruction fidelity and inference speed remains a key obstacle to practical use. To overcome these challenges, we present the first online reference-based video super-resolution framework (OnRef-VSR), which enables causal real-time reconstruction through efficient spatial and temporal feature transfers. The framework incorporates three key modules: Flow-Local Linear Transform-based Feature Enhancement (FLFE) module, Hierarchical Heterogeneous Feature Transfer (HHFT) module, and Detail Refinement with Flow-guided Local nonlinear transform (DRFL) module, which in combination realize effective integration of temporal cues and reference information. Extensive experiments demonstrate the superiority of the proposed approach. Specifically, OnRefVSR achieves a 5.2× faster speed with a 0.29 dB gain over the state-of-the-art real-time method, and its offline variant yields a 3.6× speedup and a 0.84 dB PSNR gain on the standard Vid4 benchmark, confirming both the effectiveness and practicality of OnRef-VSR for real-time high-fidelity video reconstruction.
- 17 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33 (10709908) , 41-45
SAM-iToF: SAM-Guided iToF Depth Recovery of Transparent and Mirror Objects
Indirect time-of-flight (iToF) cameras fail catastrophically when imaging transparent and reflective surfaces, as the complex superposition of light paths violates their fundamental operating principle. Existing methods either rely on overly simplistic physical models that lack adaptability or on data-driven deep learning approaches that exhibit poor generalization. In this letter, we introduce a novel, physics-informed paradigm that synergizes the strengths of both approaches. We are the first to steer a learnable iToF forward model using powerful spatial priors from the Segment Anything Model (SAM). Our framework leverages SAM’s guidance to parameterize a reflectance-transmittance mixture model, providing a physically-grounded representation of the scene. This representation is then processed by a dual-stream network, featuring a Dynamic Channel Gate (DCG) and Cross-Attention Fusion (CAF), to intelligently interpret the raw iToF signals and reconstruct a high-fidelity depth map. Extensive experiments on the MSD, Trans10 K, and ClearGrasp benchmarks demonstrate that our method not only sets a new state-of-the-art but also shows remarkable generalization from synthetic mirrors to real-world glass objects, offering a robust solution to this long-standing challenge.
- 17 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 33 (10709908) , 41-45
SAM-iToF: SAM-Guided iToF Depth Recovery of Transparent and Mirror Objects
Indirect time-of-flight (iToF) cameras fail catastrophically when imaging transparent and reflective surfaces, as the complex superposition of light paths violates their fundamental operating principle. Existing methods either rely on overly simplistic physical models that lack adaptability or on data-driven deep learning approaches that exhibit poor generalization. In this letter, we introduce a novel, physics-informed paradigm that synergizes the strengths of both approaches. We are the first to steer a learnable iToF forward model using powerful spatial priors from the Segment Anything Model (SAM). Our framework leverages SAM’s guidance to parameterize a reflectance-transmittance mixture model, providing a physically-grounded representation of the scene. This representation is then processed by a dual-stream network, featuring a Dynamic Channel Gate (DCG) and Cross-Attention Fusion (CAF), to intelligently interpret the raw iToF signals and reconstruct a high-fidelity depth map. Extensive experiments on the MSD, Trans10 K, and ClearGrasp benchmarks demonstrate that our method not only sets a new state-of-the-art but also shows remarkable generalization from synthetic mirrors to real-world glass objects, offering a robust solution to this long-standing challenge.