Jiajun Deng

RoMAN-Flow slashes inference latency while maintaining competitive performance, making AR-Normalizing Flows practical for real-world robotic manipulation.

LabDex reveals how a structured approach to task taxonomy can enhance the training and evaluation of robots in complex laboratory settings.

ThinkAfford achieves a remarkable 10.69% AP50 by decoupling affordance proposal generation from instruction grounding, setting a new standard for 3D grounding in cluttered environments.

ReTouch achieves a remarkable 18.4% improvement in success rates for dexterous manipulation by continuously refining tactile predictions in real-time.

ViCo3D achieves a groundbreaking 1.8x improvement in collaborative 3D object detection performance by merging LiDAR data with advanced visual representations from vision foundation models.

Achieving a 25.9% boost in endpoint accuracy, Next-Turn redefines how we handle speech disfluencies in real-time systems.

Speaker adaptive training combined with confidence-guided pseudo-labeling leads to substantial gains in elderly speech recognition accuracy, outperforming conventional methods.

Zero-shot adaptation can reduce word error rates by over 0.6% while achieving nearly 10 times faster real-time processing for elderly speech recognition.

Surgical robots can now perform complex tasks like knot tying and organ dissection with state-of-the-art accuracy, thanks to a new method that directly infers 3D spatial awareness from standard endoscopic images.