Spiking Neural Networks (SNN) are a class of bio-inspired neural networks that promise to bring low-power and low-latency inference to edge devices through asynchronous and sparse processing. However, being temporal models, SNNs depend heavily on expressive states to generate predictions on par with classical artificial neural networks (ANNs). These states converge only after long transient periods, and quickly decay without input data, leading to higher latency, power consumption, and lower accuracy. This work addresses this issue by initializing the state with an auxiliary ANN running at a low rate. The SNN then uses the state to generate predictions with high temporal resolution until the next initialization phase. Our hybrid ANN-SNN model thus combines the best of both worlds: It does not suffer from long state transients and state decay thanks to the ANN, and can generate predictions with high temporal resolution, low latency, and low power thanks to the SNN. We show for the task of event-based 2D and 3D human pose estimation that our method consumes 88% less power with only a 4% decrease in performance compared to its fully ANN counterparts when run at the same inference rate. Moreover, when compared to SNNs, our method achieves a 74% lower error. This research thus provides a new understanding of how ANNs and SNNs can be used to maximize their respective benefits.
相關內容
Recent semantic segmentation models accurately classify test-time examples that are similar to a training dataset distribution. However, their discriminative closed-set approach is not robust in practical data setups with distributional shifts and out-of-distribution (OOD) classes. As a result, the predicted probabilities can be very imprecise when used as confidence scores at test time. To address this, we propose a generative model for concurrent in-distribution misclassification (IDM) and OOD detection that relies on a normalizing flow framework. The proposed flow-based detector with an energy-based inputs (FlowEneDet) can extend previously deployed segmentation models without their time-consuming retraining. Our FlowEneDet results in a low-complexity architecture with marginal increase in the memory footprint. FlowEneDet achieves promising results on Cityscapes, Cityscapes-C, FishyScapes and SegmentMeIfYouCan benchmarks in IDM/OOD detection when applied to pretrained DeepLabV3+ and SegFormer semantic segmentation models.
Due to its ability to generate millions of particles, massively detailed scenes and confusing artificial illumination with reality, the version 5 of Unreal Engine promises unprecedented industrial applications. The paradigms and aims of Unreal Engine contrast with the industrial simulators typically used by the scientific community. The visual quality and performance of its rendering engine increase the opportunities, especially for industries and simulation business: where interoperability and scalability are required. The study of the following issue `` Which architecture should we implement to integrate real-world data, in an Unreal Engine 5 simulator and in a mixed-reality environment? '' offers a point of view. The topic is reexamined in an innovative and conceptual way, such as the generalization of mixedreality technologies, Internet of Things, digital twins, Big Data but providing a solution for simple and actual use cases. This paper gives a detailed analysis of the issue, at both theoretical and operational level. Then, the document goes deep into Unreal Engine's operation in order to extract the vanilla capabilities. Next, the C++ Plugin system is reviewed in details as well as the third-party library integration: pitfalls to be avoided are shown. Finally, the last chapter proposes a generic architecture, useful in large-scale industrial 3D applications, such as collaborative work or hyper-connected simulators. This document might be of interest to an Unreal Engine expert who would like to discover about server architectures. Conversely, it could be relevant for an expert in backend servers who wants to learn about Unreal Engine capabilities. This research concludes that Unreal Engine's modularity enables integration with almost any protocol. The features to integrate external real data are numerous but depend on use cases. Distributed systems for Big Data require a scalable architecture, possibly without the use of the Unreal Engine dedicated server. Environments, which require sub-second latency need to implement direct connections, bypassing any intermediate servers.
Several NP-hard problems are solved exactly using exponential-time branching strategies, whether it be branch-and-bound algorithms, or bounded search trees in fixed-parameter algorithms. The number of tractable instances that can be handled by sequential algorithms is usually small, whereas massive parallelization has been shown to significantly increase the space of instances that can be solved exactly. However, previous centralized approaches require too much communication to be efficient, whereas decentralized approaches are more efficient but have difficulty keeping track of the global state of the exploration. In this work, we propose to revisit the centralized paradigm while avoiding previous bottlenecks. In our strategy, the center has lightweight responsibilities, requires only a few bits for every communication, but is still able to keep track of the progress of every worker. In particular, the center never holds any task but is able to guarantee that a process with no work always receives the highest priority task globally. Our strategy was implemented in a generic C++ library called GemPBA, which allows a programmer to convert a sequential branching algorithm into a parallel version by changing only a few lines of code. An experimental case study on the vertex cover problem demonstrates that some of the toughest instances from the DIMACS challenge graphs that would take months to solve sequentially can be handled within two hours with our approach.
Large Language Models (LLMs), such as ChatGPT, demonstrate a strong understanding of human natural language and have been explored and applied in various fields, including reasoning, creative writing, code generation, translation, and information retrieval. By adopting LLM as the reasoning core, we introduce Autonomous GIS as an AI-powered geographic information system (GIS) that leverages the LLM's general abilities in natural language understanding, reasoning and coding for addressing spatial problems with automatic spatial data collection, analysis and visualization. We envision that autonomous GIS will need to achieve five autonomous goals including self-generating, self-organizing, self-verifying, self-executing, and self-growing. We developed a prototype system called LLM-Geo using GPT-4 API in a Python environment, demonstrating what an autonomous GIS looks like and how it delivers expected results without human intervention using two case studies. For both case studies, LLM-Geo returned accurate results, including aggregated numbers, graphs, and maps, significantly reducing manual operation time. Although still lacking several important modules such as logging and code testing, LLM-Geo demonstrates a potential path towards next-generation AI-powered GIS. We advocate for the GIScience community to dedicate more effort to the research and development of autonomous GIS, making spatial analysis easier, faster, and more accessible to a broader audience.
Drones have the potential to revolutionize power line inspection by increasing productivity, reducing inspection time, improving data quality, and eliminating the risks for human operators. Current state-of-the-art systems for power line inspection have two shortcomings: (i) control is decoupled from perception and needs accurate information about the location of the power lines and masts; (ii) collision avoidance is decoupled from the power line tracking, which results in poor tracking in the vicinity of the power masts, and, consequently, in decreased data quality for visual inspection. In this work, we propose a model predictive controller (MPC) that overcomes these limitations by tightly coupling perception and action. Our controller generates commands that maximize the visibility of the power lines while, at the same time, safely avoiding the power masts. For power line detection, we propose a lightweight learning-based detector that is trained only on synthetic data and is able to transfer zero-shot to real-world power line images. We validate our system in simulation and real-world experiments on a mock-up power line infrastructure.
A secure fingerprint recognition system must contain both a presentation attack (i.e., spoof) detection and recognition module in order to protect users against unwanted access by malicious users. Traditionally, these tasks would be carried out by two independent systems; however, recent studies have demonstrated the potential to have one unified system architecture in order to reduce the computational burdens on the system, while maintaining high accuracy. In this work, we leverage a vision transformer architecture for joint spoof detection and matching and report competitive results with state-of-the-art (SOTA) models for both a sequential system (two ViT models operating independently) and a unified architecture (a single ViT model for both tasks). ViT models are particularly well suited for this task as the ViT's global embedding encodes features useful for recognition, whereas the individual, local embeddings are useful for spoof detection. We demonstrate the capability of our unified model to achieve an average integrated matching (IM) accuracy of 98.87% across LivDet 2013 and 2015 CrossMatch sensors. This is comparable to IM accuracy of 98.95% of our sequential dual-ViT system, but with ~50% of the parameters and ~58% of the latency.
Recent advances in Transformer architectures [1] have brought remarkable improvements to visual question answering (VQA). Nevertheless, Transformer-based VQA models are usually deep and wide to guarantee good performance, so they can only run on powerful GPU servers and cannot run on capacity-restricted platforms such as mobile phones. Therefore, it is desirable to learn an elastic VQA model that supports adaptive pruning at runtime to meet the efficiency constraints of different platforms. To this end, we present the bilaterally slimmable Transformer (BST), a general framework that can be seamlessly integrated into arbitrary Transformer-based VQA models to train a single model once and obtain various slimmed submodels of different widths and depths. To verify the effectiveness and generality of this method, we integrate the proposed BST framework with three typical Transformer-based VQA approaches, namely MCAN [2], UNITER [3], and CLIP-ViL [4], and conduct extensive experiments on two commonly-used benchmark datasets. In particular, one slimmed MCAN-BST submodel achieves comparable accuracy on VQA-v2, while being 0.38x smaller in model size and having 0.27x fewer FLOPs than the reference MCAN model. The smallest MCAN-BST submodel only has 9M parameters and 0.16G FLOPs during inference, making it possible to deploy it on a mobile device with less than 60 ms latency.
This paper introduces an online model for object detection in videos designed to run in real-time on low-powered mobile and embedded devices. Our approach combines fast single-image object detection with convolutional long short term memory (LSTM) layers to create an interweaved recurrent-convolutional architecture. Additionally, we propose an efficient Bottleneck-LSTM layer that significantly reduces computational cost compared to regular LSTMs. Our network achieves temporal awareness by using Bottleneck-LSTMs to refine and propagate feature maps across frames. This approach is substantially faster than existing detection methods in video, outperforming the fastest single-frame models in model size and computational cost while attaining accuracy comparable to much more expensive single-frame models on the Imagenet VID 2015 dataset. Our model reaches a real-time inference speed of up to 15 FPS on a mobile CPU.
We introduce an effective model to overcome the problem of mode collapse when training Generative Adversarial Networks (GAN). Firstly, we propose a new generator objective that finds it better to tackle mode collapse. And, we apply an independent Autoencoders (AE) to constrain the generator and consider its reconstructed samples as "real" samples to slow down the convergence of discriminator that enables to reduce the gradient vanishing problem and stabilize the model. Secondly, from mappings between latent and data spaces provided by AE, we further regularize AE by the relative distance between the latent and data samples to explicitly prevent the generator falling into mode collapse setting. This idea comes when we find a new way to visualize the mode collapse on MNIST dataset. To the best of our knowledge, our method is the first to propose and apply successfully the relative distance of latent and data samples for stabilizing GAN. Thirdly, our proposed model, namely Generative Adversarial Autoencoder Networks (GAAN), is stable and has suffered from neither gradient vanishing nor mode collapse issues, as empirically demonstrated on synthetic, MNIST, MNIST-1K, CelebA and CIFAR-10 datasets. Experimental results show that our method can approximate well multi-modal distribution and achieve better results than state-of-the-art methods on these benchmark datasets. Our model implementation is published here: //github.com/tntrung/gaan
Recently, deep learning has achieved very promising results in visual object tracking. Deep neural networks in existing tracking methods require a lot of training data to learn a large number of parameters. However, training data is not sufficient for visual object tracking as annotations of a target object are only available in the first frame of a test sequence. In this paper, we propose to learn hierarchical features for visual object tracking by using tree structure based Recursive Neural Networks (RNN), which have fewer parameters than other deep neural networks, e.g. Convolutional Neural Networks (CNN). First, we learn RNN parameters to discriminate between the target object and background in the first frame of a test sequence. Tree structure over local patches of an exemplar region is randomly generated by using a bottom-up greedy search strategy. Given the learned RNN parameters, we create two dictionaries regarding target regions and corresponding local patches based on the learned hierarchical features from both top and leaf nodes of multiple random trees. In each of the subsequent frames, we conduct sparse dictionary coding on all candidates to select the best candidate as the new target location. In addition, we online update two dictionaries to handle appearance changes of target objects. Experimental results demonstrate that our feature learning algorithm can significantly improve tracking performance on benchmark datasets.
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