Wireless Multi-Hop Networks (WMNs) have been widely applied in many industrial monitoring applications, where typically nodes deployed on machines sample the data periodically and the data then needs to be gathered to the sink node in time. Due to the complicated industrial environment, the channel capacity between neighboring nodes differs greatly from each other. This leads to a typical dilemma of radio communication module selection, i.e., choosing the wide-bandwidth module with short communication range or the long-range module with low bandwidth. Guaranteeing both the connectivity of the network and high network throughput is not easy. To address this issue, we equip the nodes with heterogeneous radio modules, leveraging their advantages to ensure both long-distance and high-bandwidth communication capabilities. The problem of efficiently scheduling the links between the nodes with multiple heterogeneous RF modules to minimize the data gathering delay has not yet been explored in existing research. In this paper, we propose a low-latency data collection method based on Deep Reinforcement Learning, with which nodes can autonomously appropriately schedule and assign their heterogeneous RF modules to their transmission/reception targets based on the current amount of their local data as well as that of their neighbors’ data. Extensive simulations are conducted, and the results show that the proposed method can achieve 10%-25% latency reduction compared to existing multi-radio approaches.

Open-vocabulary 3D instance segmentation is cuttingedge for its ability to segment 3D instances without prede-fined categories. However, progress in 3D lags behind its 2D counterpart due to limited annotated 3D data. To ad-dress this, recent works first generate 2D open-vocabulary masks through 2D models and then merge them into 3D instances based on metrics calculated between two neigh-boring frames. In contrast to these local metrics, we pro-pose a novel metric, view consensus rate, to enhance the utilization of multi-view observations. The key insight is that two 2D masks should be deemed part of the same 3D instance if a significant number of other 2D masks from different views contain both these two masks. Using this metric as edge weight, we construct a global mask graph where each mask is a node. Through iterative clustering of masks showing high view consensus, we generate a series of clusters, each representing a distinct 3D instance. Notably, our model is training-free. Through extensive ex-periments on publicly available datasets, including Scan-Net++, ScanNet200 and MatterPo rt3D, we demonstrate that our method achieves state-of-the-art performance in open-vocabulary 3D instance segmentation. Our project page is at https://pku-epic.github.ioIMaskClustering.

In this work, we tackle the challenging task of jointly tracking hand object poses and reconstructing their shapes from depth point cloud sequences in the wild, given the initial poses at frame 0. We for the first time propose a point cloud-based hand joint tracking network, HandTrackNet, to estimate the inter-frame hand joint motion. Our HandTrackNet proposes a novel hand pose canonicalization module to ease the tracking task, yielding accurate and robust hand joint tracking. Our pipeline then reconstructs the full hand via converting the predicted hand joints into a MANO hand. For object tracking, we devise a simple yet effective module that estimates the object SDF from the first frame and performs optimization-based tracking. Finally, a joint optimization step is adopted to perform joint hand and object reasoning, which alleviates the occlusion-induced ambiguity and further refines the hand pose. During training, the whole pipeline only sees purely synthetic data, which are synthesized with sufficient variations and by depth simulation for the ease of generalization. The whole pipeline is pertinent to the generalization gaps and thus directly transferable to real in-the-wild data. We evaluate our method on two real hand object interaction datasets, e.g. HO3D and DexYCB, without any fine-tuning. Our experiments demonstrate that the proposed method significantly outperforms the previous state-of-the-art depth-based hand and object pose estimation and tracking methods, running at a frame rate of 9 FPS. We have released our code on https://github.com/PKU-EPIC/HOTrack.

The task of Vision-Language Navigation (VLN) is for an embodied agent to reach the global goal according to the instruction. Essentially, during navigation, a series of sub-goals need to be adaptively set and achieved, which is naturally a hierarchical navigation process. However, previous methods leverage a single-step planning scheme, i.e., directly performing navigation action at each step, which is unsuitable for such a hierarchical navigation process. In this paper, we propose an Adaptive Zone-aware Hierarchical Planner (AZHP) to explicitly divides the navigation process into two heterogeneous phases, i.e., sub-goal setting via zone partition/selection (high-level action) and sub-goal executing (low-level action), for hierarchical planning. Specifically, AZHP asynchronously performs two levels of action via the designed State-Switcher Module (SSM). For high-level action, we devise a Scene-aware adaptive Zone Partition (SZP) method to adaptively divide the whole navigation area into different zones on-the-fly. Then the Goal-oriented Zone Selection (GZS) method is proposed to select a proper zone for the current sub-goal. For low-level action, the agent conducts navigation-decision multi-steps in the selected zone. Moreover, we design a Hierarchical RL (HRL) strategy and auxiliary losses with curriculum learning to train the AZHP framework, which provides effective supervision signals for each stage. Extensive experiments demonstrate the superiority of our proposed method, which achieves state-of-the-art performance on three VLN benchmarks (REVERIE, SOON, R2R).

As a popular geometric representation, point clouds have attracted much attention in 3D vision, leading to many applications in autonomous driving and robotics. One important yet unsolved issue for learning on point cloud is that point clouds of the same object can have significant geometric variations if generated using different procedures or captured using different sensors. These inconsistencies induce domain gaps such that neural networks trained on one domain may fail to generalize on others. A typical technique to reduce the domain gap is to perform adversarial training so that point clouds in the feature space can align. However, adversarial training is easy to fall into degenerated local minima, resulting in negative adaptation gains. Here we propose a simple yet effective method for unsupervised domain adaptation on point clouds by employing a self-supervised task of learning geometry-aware implicits, which plays two critical roles in one shot. First, the geometric information in the point clouds is preserved through the implicit representations for downstream tasks. More importantly, the domain-specific variations can be effectively learned away in the implicit space. We also propose an adaptive strategy to compute unsigned distance fields for arbitrary point clouds due to the lack of shape models in practice. When combined with a task loss, the proposed outperforms state-of-the-art unsupervised domain adaptation methods that rely on adversarial domain alignment and more complicated self-supervised tasks. Our method is evaluated on both PointDA-10 and GraspNet datasets. Code and data are available at: https://github.com/Jhonve/ImplicitPCDA.

There exist the generic concepts and semantic relationships in the description of procedural knowledge and strategic knowledge in contract management. In order that contract information could be understood and processed by machine automatically, more semantic descriptions are needed. Ontology has become a promising technology to express semantics. At present, Web ontology languages have been standardized. However, building domain ontology based on OWL (Ontology Web Language) just begins and has a long way to go. In building domain ontology, people should explore the rationality in expressing domain concepts and relations between the concepts, considering the capabilities of OWL (not the capability of natural languages). In this paper, we put forward a framework of contract ontology after considering capabilities to express semantic relations provided by OWL. Then we use Protege-2000 to build OWL representations of the contract ontology and make some analyses. Our exploration might not standardize contract ontology. We explore with the intention to help people better understand the rationality in expressing concepts and relations between the concepts in contract management in order that OWL contract ontology and other domain ontology are improved and standardized as soon as possible

Constructing phylogenetic trees in the study of the evolutionary history of a group organisms is an extremely challenging problem in computational biology. The problem becomes intractable with growing number of organisms. In this paper, we design and implement an efficient parallel solver (ExactMP) using a parsimony based approach for solving this problem. We create a testbed consisting of eighteen datasets of varying size (up to 27 taxa) and difficulty level (easy to hard), containing real (Eukaryotes, Metazoan, and rbcL) and randomly-generated synthetic genome sequences. We demonstrate our ExactMP Solver against this testbed and achieve a parallel speedup of up to 7.26 with 8 processors using an 8-way symmetric multiprocessor. The main contributions of this work are: (1) an efficient parallel solver ExactMP for the problem of phylogenetic tree reconstruction using maximum parsimony, (2) a new upper bounding methodology for this problem using heuristic and randomization techniques, and (3) a highly optimized branch and bound algorithm for this problem.