Xiaotong Liu

Interrupted voice agents often hallucinate what they have already said because text generation outpaces audio playback—a failure mode solved by forcing the model to listen to its own real-time voice output.

Existing staypoint detection algorithms falter in noisy environments, but new unsupervised methods show promise for substantial improvements.

You can now get accurate time-series forecasts with feature-level interpretability, enabling simpler and more efficient early warning systems.

Fisheye cameras in robotics offer superior scene generalization, but only if you train them with enough environmental diversity to avoid overfitting.