Humans perform everyday tasks using a combination of locomotion and manipulation skills. Building a system that can handle both skills is essential to creating virtual humans. We present a physically-simulated human capable of solving box rearrangement tasks, which requires a combination of both skills. We propose a hierarchical control architecture, where each level solves the task at a different level of abstraction, and the result is a physics-based simulated virtual human capable of rearranging boxes in a cluttered environment. The control architecture integrates a planner, diffusion models, and physics-based motion imitation of sparse motion clips using deep reinforcement learning. Boxes can vary in size, weight, shape, and placement height. Code and trained control policies are provided.
相關內容
We propose a location-adaptive self-normalization (SN) based test for change points in time series. The SN technique has been extensively used in change-point detection for its capability to avoid direct estimation of nuisance parameters. However, we find that the power of the SN-based test is susceptible to the location of the break and may suffer from a severe power loss, especially when the change occurs at the early or late stage of the sequence. This phenomenon is essentially caused by the unbalance of the data used before and after the change point when one is building a test statistic based on the cumulative sum (CUSUM) process. Hence, we consider leaving out the samples far away from the potential locations of change points and propose an optimal data selection scheme. Based on this scheme, a new SN-based test statistic adaptive to the locations of breaks is established. The new test can significantly improve the power of the existing SN-based tests while maintaining a satisfactory size. It is a unified treatment that can be readily extended to tests for general quantities of interest, such as the median and the model parameters. The derived optimal subsample selection strategy is not specific to the SN-based tests but is applicable to any method that relies on the CUSUM process, which may provide new insights in the area for future research.
Clinically deployed segmentation models are known to fail on data outside of their training distribution. As these models perform well on most cases, it is imperative to detect out-of-distribution (OOD) images at inference to protect against automation bias. This work applies the Mahalanobis distance post hoc to the bottleneck features of a Swin UNETR model that segments the liver on T1-weighted magnetic resonance imaging. By reducing the dimensions of the bottleneck features with principal component analysis, OOD images were detected with high performance and minimal computational load.
Spiking Neural Networks (SNNs) are promising energy-efficient models for neuromorphic computing. For training the non-differentiable SNN models, the backpropagation through time (BPTT) with surrogate gradients (SG) method has achieved high performance. However, this method suffers from considerable memory cost and training time during training. In this paper, we propose the Spatial Learning Through Time (SLTT) method that can achieve high performance while greatly improving training efficiency compared with BPTT. First, we show that the backpropagation of SNNs through the temporal domain contributes just a little to the final calculated gradients. Thus, we propose to ignore the unimportant routes in the computational graph during backpropagation. The proposed method reduces the number of scalar multiplications and achieves a small memory occupation that is independent of the total time steps. Furthermore, we propose a variant of SLTT, called SLTT-K, that allows backpropagation only at K time steps, then the required number of scalar multiplications is further reduced and is independent of the total time steps. Experiments on both static and neuromorphic datasets demonstrate superior training efficiency and performance of our SLTT. In particular, our method achieves state-of-the-art accuracy on ImageNet, while the memory cost and training time are reduced by more than 70% and 50%, respectively, compared with BPTT.
Massive emerging applications are driving demand for the ubiquitous deployment of computing power today. This trend not only spurs the recent popularity of the \emph{Computing and Network Convergence} (CNC), but also introduces an urgent need for the intelligentization of a management platform to coordinate changing resources and tasks in the CNC. Therefore, in this article, we present the concept of an intelligence-endogenous management platform for CNCs called \emph{CNC brain} based on artificial intelligence technologies. It aims at efficiently and automatically matching the supply and demand with high heterogeneity in a CNC via four key building blocks, i.e., perception, scheduling, adaptation, and governance, throughout the CNC's life cycle. Their functionalities, goals, and challenges are presented. To examine the effectiveness of the proposed concept and framework, we also implement a prototype for the CNC brain based on a deep reinforcement learning technology. Also, it is evaluated on a CNC testbed that integrates two open-source and popular frameworks (OpenFaas and Kubernetes) and a real-world business dataset provided by Microsoft Azure. The evaluation results prove the proposed method's effectiveness in terms of resource utilization and performance. Finally, we highlight the future research directions of the CNC brain.
Embedded devices are increasingly present in our everyday life. They often process critical information, and hence, rely on cryptographic protocols to achieve security. However, embedded devices remain vulnerable to attackers seeking to hijack their operation and extract sensitive information by exploiting side channels and code reuse. Code-Reuse Attacks (CRAs) can steer the execution of a program to malicious outcomes, altering existing on-board code without direct access to the device memory. Moreover, Side-Channel Attacks (SCAs) may reveal secret information to the attacker based on mere observation of the device. Thwarting CRAs and SCAs against embedded devices is challenging because embedded devices are often resource constrained. Fine-grained code diversification hinders CRAs by introducing uncertainty to the binary code; while software mechanisms can thwart timing or power SCAs. The resilience to either attack may come at the price of the overall efficiency. Moreover, a unified approach that preserves these mitigations against both CRAs and SCAs is not available. In this paper, we propose a novel Secure Diversity by Construction (SecDivCon) approach that tackles this challenge. SecDivCon is a combinatorial compiler-based approach that combines software diversification against CRAs with software mitigations against SCAs. SecDivCon restricts the performance overhead introduced by the generated code that thwarts the attacks and hence, offers a secure-by-design approach enabling control over the performance-security trade-off. Our experiments, using 16 benchmark programs, show that SCA-aware diversification is effective against CRAs, while preserving SCA mitigation properties at a low, controllable overhead. Given the combinatorial nature of our approach, SecDivCon is suitable for small, performance-critical functions that are sensitive to SCAs.
In many high-frequency simulation workflows, eigenvalue tracking along a parameter variation is necessary. This can become computationally prohibitive when repeated time-consuming eigenvalue problems must be solved. Therefore, we employ a reduced basis approximation to bring down the computational costs. It is based on the greedy strategy from Horger et al. 2017 which considers multiple eigenvalues for elliptic eigenvalue problems. We extend this algorithm to deal with parameter-dependent domains and the Maxwell eigenvalue problem. In this setting, the reduced basis may contain spurious eigenmodes, which require special treatment. We demonstrate our algorithm in an eigenvalue tracking application for an eigenmode classification.
Answering questions that require reading texts in an image is challenging for current models. One key difficulty of this task is that rare, polysemous, and ambiguous words frequently appear in images, e.g., names of places, products, and sports teams. To overcome this difficulty, only resorting to pre-trained word embedding models is far from enough. A desired model should utilize the rich information in multiple modalities of the image to help understand the meaning of scene texts, e.g., the prominent text on a bottle is most likely to be the brand. Following this idea, we propose a novel VQA approach, Multi-Modal Graph Neural Network (MM-GNN). It first represents an image as a graph consisting of three sub-graphs, depicting visual, semantic, and numeric modalities respectively. Then, we introduce three aggregators which guide the message passing from one graph to another to utilize the contexts in various modalities, so as to refine the features of nodes. The updated nodes have better features for the downstream question answering module. Experimental evaluations show that our MM-GNN represents the scene texts better and obviously facilitates the performances on two VQA tasks that require reading scene texts.
Knowledge graphs are important resources for many artificial intelligence tasks but often suffer from incompleteness. In this work, we propose to use pre-trained language models for knowledge graph completion. We treat triples in knowledge graphs as textual sequences and propose a novel framework named Knowledge Graph Bidirectional Encoder Representations from Transformer (KG-BERT) to model these triples. Our method takes entity and relation descriptions of a triple as input and computes scoring function of the triple with the KG-BERT language model. Experimental results on multiple benchmark knowledge graphs show that our method can achieve state-of-the-art performance in triple classification, link prediction and relation prediction tasks.
Many tasks in natural language processing can be viewed as multi-label classification problems. However, most of the existing models are trained with the standard cross-entropy loss function and use a fixed prediction policy (e.g., a threshold of 0.5) for all the labels, which completely ignores the complexity and dependencies among different labels. In this paper, we propose a meta-learning method to capture these complex label dependencies. More specifically, our method utilizes a meta-learner to jointly learn the training policies and prediction policies for different labels. The training policies are then used to train the classifier with the cross-entropy loss function, and the prediction policies are further implemented for prediction. Experimental results on fine-grained entity typing and text classification demonstrate that our proposed method can obtain more accurate multi-label classification results.
Dynamic programming (DP) solves a variety of structured combinatorial problems by iteratively breaking them down into smaller subproblems. In spite of their versatility, DP algorithms are usually non-differentiable, which hampers their use as a layer in neural networks trained by backpropagation. To address this issue, we propose to smooth the max operator in the dynamic programming recursion, using a strongly convex regularizer. This allows to relax both the optimal value and solution of the original combinatorial problem, and turns a broad class of DP algorithms into differentiable operators. Theoretically, we provide a new probabilistic perspective on backpropagating through these DP operators, and relate them to inference in graphical models. We derive two particular instantiations of our framework, a smoothed Viterbi algorithm for sequence prediction and a smoothed DTW algorithm for time-series alignment. We showcase these instantiations on two structured prediction tasks and on structured and sparse attention for neural machine translation.
小貼士
登錄享
相關主題
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191