MOTIVATION: Structure-based drug design (SBDD) has advanced with deep generative models, but bridging the gap between continuous atomic coordinates and discrete atom types remains a challenge. Current approaches, such as diffusion and flow matching models, often fail to unify these heterogeneous modalities, relying on separate strategies or ill-fitting Euclidean metrics for discrete variables. This lack of a consistent framework limits generative models' ability to capture the geometric and chemical structure of protein-ligand complexes. RESULTS: We present MolPIF, a parameter interpolation flow mechanism designed to unify the generation of continuous and discrete molecular variables. Unlike traditional flow models that operate in sample space, MolPIF interpolates between distributions in the parameter space, theoretically recovering Wasserstein-2 optimal transport for continuous coordinates and establishing Fisher-Rao geodesics for discrete atom types. We further incorporate a geometry-enhanced learning strategy to improve the capture of atomic contexts. Extensive evaluations on the CrossDocked2020 dataset demonstrate that MolPIF outperforms baselines in binding affinity, chemical validity, geometric fidelity, and chemical space coverage. Additionally, MolPIF exhibits versatility in lead optimization and offers flexible prior distribution selection (such as Laplace), establishing a robust paradigm for SBDD. AVAILABILITY AND IMPLEMENTATION: Source code is freely available at https://github.com/BLEACH366/MolPIF.

Deep learning-based feature matching has showcased great superiority for point cloud registration. While coarse-to-fine matching architectures are prevalent, they typically perform sparse and geometrically inconsistent coarse matching. This forces the subsequent fine matching to rely on computationally expensive optimal transport and hypothesis-and-selection procedures to resolve inconsistencies, leading to inefficiency and poor scalability for large-scale real-time applications. In this paper, we design a consistency-aware spot-guided Transformer (CAST) to enhance the coarse matching by explicitly utilizing geometric consistency via two key sparse attention mechanisms. First, our consistency-aware self-attention selectively computes intra-point-cloud attention to a sparse subset of points with globally consistent correspondences, enabling other points to derive discriminative features through their relationships with these anchors while propagating global consistency for robust correspondence reasoning. Second, our spot-guided cross-attention restricts cross-point-cloud attention to dynamically defined "spots"-the union of correspondence neighborhoods of a query's neighbors in the other point cloud, which are most likely to cover the true correspondence of the query ensured by local consistency, eliminating interference from similar but irrelevant regions. Furthermore, we design a lightweight local attention-based fine matching module to precisely predict dense correspondences and estimate the transformation. Extensive experiments on both outdoor LiDAR datasets and indoor RGB-D camera datasets demonstrate that our method achieves state-of-the-art accuracy, efficiency, and robustness. Besides, our method showcases superior generalization ability on our newly constructed challenging relocalization and loop closing benchmarks in unseen domains.

Collaborative edge-device inference is a promising way to empower resource-constrained mobile devices to execute deep neural network (DNN)-based applications with heavy computational workloads. In particular, a DNN model is partitioned into two parts that are executed on the mobile device and the edge server, respectively. However, offline model partitioning methods suffer from poor adaptability to real computing environments, while online methods have the problem of delayed feedback. In addition, model partitioning inevitably incurs large transmission overheads of DNN's intermediate data. To tackle these challenges, we propose a collaborative edge-device inference optimization algorithm (JPCA) with Joint DNN Partitioning, data Compression, and resource Allocation. Our goal is to maximize the inference accuracy of all tasks while satisfying inference latency and energy requirements in a dynamically changing environment. In JPCA, the joint selection of model partition point and data quantization bit-width is first extracted from the original problem and we propose an improved deep reinforcement learning (DRL)-based algorithm to learn joint decisions. Optimal schemes under different bandwidth conditions are recorded, enabling mobile devices to adjust joint decisions in response to significant bandwidth changes. Furthermore, we design a dynamic edge resource allocation algorithm that makes edge resource allocation decisions for tasks arriving in real time, thereby accelerating DNN inference. The results of the testbed experiments affirm the effectiveness of our proposed algorithms in terms of inference accuracy.

With the increase in computing speed and network bandwidth in data centers, microsecond-level tail latency has become a key metric for internet-based online services. However, the tail latency of messages in data centers is mainly determined by queuing delay, which is usually much larger than pure message transmission time. Existing traffic scheduling mechanisms fail to effectively coordinate end-side and in-network resources, leading to head-of-line (HOL) blocking for microsecond-level messages at both ends and switches, which severely impacts the tail latency. To address this issue, this paper proposes a network-wide Global Priority based Multi-Path message scheduling mechanism GPMP. It assigns the highest priority to microsecond-level messages at both ends and swtiches to ensure such messages can quickly acquire resources and complete quickly. Furthermore, GPMP also optimizes the transmission of low-priority large messages by introducing a multi-path method, which effectively reduces transmission time and improves the bandwidth utilization. Extensive simulation results show that GPMP not only meets the strict latency requirements of microsecond-level messages, but also significantly reduces the completion time of long messages, leading to an overall improvement in bandwidth utilization.

Deep learning-based feature matching has shown great superiority for point cloud registration in the absence of pose priors. Although coarse-to-fine matching approaches are prevalent, the coarse matching of existing methods is typically sparse and loose without consideration of geometric consistency, which makes the subsequent fine matching rely on ineffective optimal transport and hypothesis-and-selection methods for consistency. Therefore, these methods are neither efficient nor scalable for real-time applications such as odometry in robotics. To address these issues, we design a consistency-aware spot-guided Transformer (CAST), which incorporates a spot-guided cross-attention module to avoid interfering with irrelevant areas, and a consistency-aware self-attention module to enhance matching capabilities with geometrically consistent correspondences. Furthermore, a lightweight fine matching module for both sparse keypoints and dense features can estimate the transformation accurately. Extensive experiments on both outdoor LiDAR point cloud datasets and indoor RGBD point cloud datasets demonstrate that our method achieves state-of-the-art accuracy, efficiency, and robustness.