SearcharxivSearch

arXiv subjects

Yao Chang

Publications and source records attributed to Yao Chang.

2 recordsLinked to original sources

Evolution of dipole-dipole dynamics in cold ammonia collisions

Cold polar molecules offer fascinating prospects for ultracold chemistry and quantum physics, including new platforms for quantum simulation or computation. However, their inherent collision properties remain largely unknown. It has proven extremely hard to experimentally probe collisions between two dipolar molecules at sufficiently low energies and high precision, as it appears fundamentally impossible to merge two beams of molecules with significant dipole moments. Here we report measurements of state-to-state cross sections for collisions between ammonia isotopologues at energies between 0.3 and 100 cm$^{-1}$ using a novel beam merger. We experimentally observed a local maximum in the cross sections that indicates the effective dipole moments can switch off at low collision energies. Scattering calculations reproduced this maximum in good agreement and explained the observed scaling with the parity splitting energies in the molecular energy level structures. Measurements of the correlated energy transfer in both collision partners yielded direct evidence of the suppression of the dipole-dipole interaction at low energies. Our results demonstrate how collisions between an important class of polar molecules evolve from the high temperature limit towards the ultracold regime in a counterintuitive way, have major consequences for the feasibility of future experiments and the interpretation of previous work, and offer distinctive opportunities to control cold molecular collisions with external fields.

physics.atom-ph

TransClaw U-Net: Claw U-Net with Transformers for Medical Image Segmentation

In recent years, computer-aided diagnosis has become an increasingly popular topic. Methods based on convolutional neural networks have achieved good performance in medical image segmentation and classification. Due to the limitations of the convolution operation, the long-term spatial features are often not accurately obtained. Hence, we propose a TransClaw U-Net network structure, which combines the convolution operation with the transformer operation in the encoding part. The convolution part is applied for extracting the shallow spatial features to facilitate the recovery of the image resolution after upsampling. The transformer part is used to encode the patches, and the self-attention mechanism is used to obtain global information between sequences. The decoding part retains the bottom upsampling structure for better detail segmentation performance. The experimental results on Synapse Multi-organ Segmentation Datasets show that the performance of TransClaw U-Net is better than other network structures. The ablation experiments also prove the generalization performance of TransClaw U-Net.

cs.CV