arXiv subjects

Weiwei Zhou

Publications and source records attributed to Weiwei Zhou.

11 recordsLinked to original sources

arXiv2026

CIGTSurv: Clinical Information Guided Tri-modal Survival Prediction with Local Prototype Association and Global Feature Alignment

Multimodal learning has significantly advanced survival prediction by integrating pathology images with genomic data. However, clinical information, despite its critical role in reflecting a patient' s overall health, remains underutilized due to its discrete, sparse, and low-dimensional nature. Furthermore, the inherent heterogeneity across these modalities pose significant challenges in modeling cross-modal interactions. In this paper, we propose CIGTSurv, a Clinical Information Guided Tri-modal framework for Survival prediction. Specifically, we first design a holistic text template and use pretrained foundation models to transform clinical tabular data into high-dimensional tokenized embeddings. Using clinical information as an anchor, we then introduce a dual-level interaction mechanism: 1) a local prototype association (LPA) module based on cross-attention to explicitly learn token-level correspondences between different modalities, and 2) a global feature alignment (GFA) loss based on Maximum Mean Discrepancy (MMD) to implicitly enhance cross-modal distribution consistency. Extensive experiments on five TCGA cancer cohorts demonstrate that CIGTSurv achieves state-of-the-art (SOTA) survival prediction performance. Our source code is publicly available at https://github.com/Daijing-ai/CIGT-Surv.git.

cs.CV↗

arXiv2025

Towards Cellular-Scale Interpretability in Pathology Foundation Models for Biomarker Assessment

Molecular biomarker testing in pathology is often costly and tissue-consuming, limiting scalable clinical deployment. Artificial intelligence applied to hematoxylin and eosin (HE)-stained histology could enable rapid biomarker screening, but clinical translation requires models that are both accurate and interpretable. Here we introduce Hireca, a biomarker-focused pathology foundation model pretrained on more than 80,000 whole-slide images spanning 38 organ types from three medical centers, together with CytoMap, an interpretability module that localizes cellular-scale evidence underlying predictions. Across 10 biomarker tasks encompassing morphological, molecular, genetic, and spatial-transcriptomic-proxy readouts, Hireca ranked first in five tasks and outperformed comparable models overall. In evaluation by eight pathologists from two countries, CytoMap was consistently preferred over alternative visualization approaches and revealed error patterns in difficult cases. These results position Hireca and CytoMap as a transparent framework for clinically reviewable biomarker assessment directly from routine HE histology.

cs.CV↗

arXiv2025

Emotion Recognition with CLIP and Sequential Learning

Human emotion recognition plays a crucial role in facilitating seamless interactions between humans and computers. In this paper, we present our innovative methodology for tackling the Valence-Arousal (VA) Estimation Challenge, the Expression Recognition Challenge, and the Action Unit (AU) Detection Challenge, all within the framework of the 8th Workshop and Competition on Affective Behavior Analysis in-the-wild (ABAW). Our approach introduces a novel framework aimed at enhancing continuous emotion recognition. This is achieved by fine-tuning the CLIP model with the aff-wild2 dataset, which provides annotated expression labels. The result is a fine-tuned model that serves as an efficient visual feature extractor, significantly improving its robustness. To further boost the performance of continuous emotion recognition, we incorporate Temporal Convolutional Network (TCN) modules alongside Transformer Encoder modules into our system architecture. The integration of these advanced components allows our model to outperform baseline performance, demonstrating its ability to recognize human emotions with greater accuracy and efficiency.

cs.CV↗

arXiv2024

Boosting Continuous Emotion Recognition with Self-Pretraining using Masked Autoencoders, Temporal Convolutional Networks, and Transformers

Human emotion recognition holds a pivotal role in facilitating seamless human-computer interaction. This paper delineates our methodology in tackling the Valence-Arousal (VA) Estimation Challenge, Expression (Expr) Classification Challenge, and Action Unit (AU) Detection Challenge within the ambit of the 6th Workshop and Competition on Affective Behavior Analysis in-the-wild (ABAW). Our study advocates a novel approach aimed at refining continuous emotion recognition. We achieve this by initially harnessing pre-training with Masked Autoencoders (MAE) on facial datasets, followed by fine-tuning on the aff-wild2 dataset annotated with expression (Expr) labels. The pre-trained model serves as an adept visual feature extractor, thereby enhancing the model's robustness. Furthermore, we bolster the performance of continuous emotion recognition by integrating Temporal Convolutional Network (TCN) modules and Transformer Encoder modules into our framework.

cs.CV↗

arXiv2023

Leveraging TCN and Transformer for effective visual-audio fusion in continuous emotion recognition

Human emotion recognition plays an important role in human-computer interaction. In this paper, we present our approach to the Valence-Arousal (VA) Estimation Challenge, Expression (Expr) Classification Challenge, and Action Unit (AU) Detection Challenge of the 5th Workshop and Competition on Affective Behavior Analysis in-the-wild (ABAW). Specifically, we propose a novel multi-modal fusion model that leverages Temporal Convolutional Networks (TCN) and Transformer to enhance the performance of continuous emotion recognition. Our model aims to effectively integrate visual and audio information for improved accuracy in recognizing emotions. Our model outperforms the baseline and ranks 3 in the Expression Classification challenge.

cs.CV↗

arXiv2014

Model discrimination for dephasing two-level systems

The problem of model discriminability and parameter identifiability for dephasing two-level systems subject to Hamiltonian control is studied. Analytic solutions of the Bloch equations are used to derive explicit expressions for observables as functions of time for different models. This information is used to give criteria for model discrimination and parameter estimation based on simple experimental paradigms.

quant-ph↗

arXiv2013

High-throughput Imaging and Spectroscopy of Individual Carbon Nanotubes in Devices with Light Microscopy

Two paramount challenges in carbon nanotube research are achieving chirality-controlled synthesis and understanding chirality-dependent device physics. High-throughput and in-situ chirality and electronic structural characterization of individual carbon nanotubes is crucial for addressing these challenges. Optical imaging and spectroscopy has unparalleled throughput and specificity, but its realization for single nanotubes on substrates or in devices has long been an outstanding challenge. Here we demonstrate video-rate imaging and in-situ spectroscopy of individual carbon nanotubes on various substrates and in functional devices using a novel high-contrast polarization-based optical microscopy. Our technique enables the complete chirality profiling of hundreds of as-grown carbon nanotubes. In addition, we in-situ monitor nanotube electronic structure in active field-effect devices, and observe that high-order nanotube optical resonances are dramatically broadened by electrostatic doping. This unexpected behaviour points to strong interband electron-electron scattering processes that can dominate ultrafast dynamics of excited states in carbon nanotubes.

cond-mat.mes-hall↗

arXiv2012

Open quantum system identification

Engineering quantum systems offers great opportunities both technologically and scientifically for communication, computation, and simulation. The construction and operation of large scale quantum information devices presents a grand challenge and a major issue is the effective control of coherent dynamics. This is often in the presence of decoherence which further complicates the task of determining the behaviour of the system. Here, we show how to determine open system Markovian dynamics of a quantum system with restricted initialisation and partial output state information.

quant-ph↗

arXiv2012

Trajectory-constrained optimal local time-continuous waveform controls for state transitions in $N$-level quantum systems

Based on a parametrization of pure quantum states we explicitly construct a sequence of (at most) $4N-5$ local time-continuous waveform controls to achieve a specified state transition for $N$-level quantum systems when sufficient controls of the Hamiltonian are available. The control magnitudes are further optimized in terms of a time-energy performance, which is a generalization of the time performance index. Trajectory-constrained optimal local time-continuous waveform controls, including both local sine-waveforms and $n^{\rm th}$-order-polynomial waveform controls are obtained in terms of time-energy performance. It is demonstrated that constrained optimal local $n^{\rm th}$-order-polynomial waveform controls approach constrained optimal bang-bang controls when $n\rightarrow\infty$.

quant-ph↗

arXiv2011

Bang-Bang Control Design for Quantum State Transfer based on Hyperspherical Coordinates and Optimal Time-energy Control

We present a constructive control scheme for solving quantum state engineering problems based on a parametrization of the state vector in terms of complex hyperspherical coordinates. Unlike many control schemes based on factorization of unitary operators the scheme gives explicit expressions for all the Euler angles in terms of the hyperspherical coordinates of the initial and final state. The factorization, when applicable, has added benefits that phase rotations can be combined and performed concurrently. The control procedure can be realized using a simple bang-bang or square-wave-function controls. Optimal time-energy control is considered to find the optimal control amplitudes. The extension of the scheme to implement unitary operators is also discussed.

quant-ph↗

arXiv2010

Effective Strategies for Identifying Model Parameters for Open Quantum Systems

The problem of identifiability of model parameters for open quantum systems is considered by investigating two-level dephasing systems. We discuss under which conditions full information about the Hamiltonian and dephasing parameters can be obtained. Using simulated experiments several different strategies for extracting model parameters from limited and noisy data are compared.

quant-ph↗