Soft-tissue motion introduces significant challenges in robotic teleoperation, especially in medical scenarios where precise target tracking is critical. Latency across sensing, computation, and actuation chains leads to degraded tracking performance, particularly around high-acceleration segments and trajectory inflection points. This study investigates machine learning-based predictive compensation for latency mitigation in soft-tissue tracking. Three models—autoregressive (AR), long short-term memory (LSTM), and temporal convolutional network (TCN)—were implemented and evaluated on both synthetic and real datasets. By aligning the prediction horizon with the end-to-end system delay, we demonstrate that prediction-based compensation significantly reduces tracking errors. Among the models, TCN achieved superior robustness and accuracy on complex motion patterns, particularly in multi-step prediction tasks, and exhibited better latency–horizon compatibility. The results suggest that TCN is a promising candidate for real-time latency compensation in teleoperated robotic systems involving dynamic soft-tissue interaction.

Robust stereo disparity estimation plays a critical role in minimally invasive surgery, where dynamic soft tissues, specular reflections, and data scarcity pose major challenges to traditional end-to-end deep learning and deformable model-based methods. In this paper, we propose a novel disparity estimation framework that leverages a pretrained StyleGAN generator to represent the disparity manifold of Minimally Invasive Surgery (MIS) scenes and reformulates the stereo matching task as a latent-space optimization problem. Specifically, given a stereo pair, we search for the optimal latent vector in the intermediate latent space of StyleGAN, such that the photometric reconstruction loss between the stereo images is minimized while regularizing the latent code to remain within the generator’s high-confidence region. Unlike existing encoder-based embedding methods, our approach directly exploits the geometry of the learned latent space and enforces both photometric consistency and manifold prior during inference, without the need for additional training or supervision. Extensive experiments on stereo-endoscopic videos demonstrate that our method achieves high-fidelity and robust disparity estimation across varying lighting, occlusion, and tissue dynamics, outperforming Thin Plate Spline (TPS)-based and linear representation baselines. This work bridges generative modeling and 3D perception by enabling efficient, training-free disparity recovery from pre-trained generative models with reduced inference latency.