Models trained with primitive awareness achieve task transfer with only three demonstrations
Explore technology by industry and role
Newest additions first. Explore past and current notes by industry and role.
All notes
Newest additions first. Explore past and current notes by industry and role.
A hierarchical framework with six agents collaboratively enables hardware-algorithm mapping and adaptive debugging
Separates illumination-sensitive appearance tokens from pose-relevant structure tokens, enabling reliable pose estimation under varying lighting conditions
A structured demonstration strategy that progressively increases task complexity enables more efficient and stable learning in VLA models.
A quantitative metric for timing uncertainty is developed to identify Delta Scuti stars suitable for spacecraft navigation.
By leveraging the unexecuted tail of the previous action chunk, SEAM enables smooth transitions between action chunks without policy retraining.
Integrated analysis of 27 genomic clusters and 50 hub genes reveals coordinated regulatory networks among heat shock-responsive genes
MiLSD proposes a new architecture enabling accurate line segment detection within ultra-low memory constraints.
Systematically evaluates how orbital uncertainties and velocity modeling errors propagate into LISA's high-frequency response
Failure-driven analysis reveals intrinsic initialization as the dominant failure factor in fisheye calibration
Modular AI acceleration and system-level optimization achieve significant performance gains
Enhances geometric guidance by integrating local texture enhancement and structured geometric region alignment in pose regression
A unified charge-mediated theory connects microscopic and macroscopic switching dynamics across semiconductor elements
A new method using quantum Hall plateau width to assess disorder in semiconductor heterostructures is proposed.
A lightweight, unsupervised framework combining self-calibrated learning and Retinex physical priors for low-light enhancement