Intel (Germany)

Multi-access edge computing (MEC) is an emerging ecosystem, which aims at converging telecommunication and IT services, providing a cloud computing platform at the edge of the radio access network. MEC offers storage and computational resources at the edge, reducing latency for mobile end users and utilizing more efficiently the mobile backhaul and core networks. This paper introduces a survey on MEC and focuses on the fundamental key enabling technologies. It elaborates MEC orchestration considering both individual services and a network of MEC platforms supporting mobility, bringing light into the different orchestration deployment options. In addition, this paper analyzes the MEC reference architecture and main deployment scenarios, which offer multitenancy support for application developers, content providers, and third parties. Finally, this paper overviews the current standardization activities and elaborates further on open research challenges.

We show that the YOLOv4 object detection neural network based on the CSP approach, scales both up and down and is applicable to small and large networks while maintaining optimal speed and accuracy. We propose a network scaling approach that modifies not only the depth, width, resolution, but also structure of the network. YOLOv4-large model achieves state-of-the-art results: 55.5% AP (73.4% AP <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">50</inf> ) for the MS COCO dataset at a speed of ~ 16 FPS on Tesla V100, while with the test time augmentation, YOLOv4-large achieves 56.0% AP (73.3 AP <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">50</inf> ). To the best of our knowledge, this is currently the highest accuracy on the COCO dataset among any published work. The YOLOv4-tiny model achieves 22.0% AP (42.0% AP <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">50</inf> ) at a speed of ~443 FPS on RTX 2080Ti, while by using TensorRT, batch size = 4 and FP16-precision the YOLOv4-tiny achieves 1774 FPS.

Legged robots pose one of the greatest challenges in robotics. Dynamic and agile maneuvers of animals cannot be imitated by existing methods that are crafted by humans. A compelling alternative is reinforcement learning, which requires minimal craftsmanship and promotes the natural evolution of a control policy. However, so far, reinforcement learning research for legged robots is mainly limited to simulation, and only few and comparably simple examples have been deployed on real systems. The primary reason is that training with real robots, particularly with dynamically balancing systems, is complicated and expensive. In the present work, we introduce a method for training a neural network policy in simulation and transferring it to a state-of-the-art legged system, thereby leveraging fast, automated, and cost-effective data generation schemes. The approach is applied to the ANYmal robot, a sophisticated medium-dog-sized quadrupedal system. Using policies trained in simulation, the quadrupedal machine achieves locomotion skills that go beyond what had been achieved with prior methods: ANYmal is capable of precisely and energy-efficiently following high-level body velocity commands, running faster than before, and recovering from falling even in complex configurations.

The success of monocular depth estimation relies on large and diverse training sets. Due to the challenges associated with acquiring dense ground-truth depth across different environments at scale, a number of datasets with distinct characteristics and biases have emerged. We develop tools that enable mixing multiple datasets during training, even if their annotations are incompatible. In particular, we propose a robust training objective that is invariant to changes in depth range and scale, advocate the use of principled multi-objective learning to combine data from different sources, and highlight the importance of pretraining encoders on auxiliary tasks. Armed with these tools, we experiment with five diverse training datasets, including a new, massive data source: 3D films. To demonstrate the generalization power of our approach we use zero-shot cross-dataset transfer, i.e. we evaluate on datasets that were not seen during training. The experiments confirm that mixing data from complementary sources greatly improves monocular depth estimation. Our approach clearly outperforms competing methods across diverse datasets, setting a new state of the art for monocular depth estimation.

The success of monocular depth estimation relies on large and diverse training sets. Due to the challenges associated with acquiring dense ground-truth depth across different environments at scale, a number of datasets with distinct characteristics and biases have emerged. We develop tools that enable mixing multiple datasets during training, even if their annotations are incompatible. In particular, we propose a robust training objective that is invariant to changes in depth range and scale, advocate the use of principled multi-objective learning to combine data from different sources, and highlight the importance of pretraining encoders on auxiliary tasks. Armed with these tools, we experiment with six diverse training datasets, including a new, massive data source: 3D films. To demonstrate the generalization power of our approach we use zero-shot cross-dataset transfer, i.e. we evaluate on datasets that were not seen during training. The experiments confirm that mixing data from complementary sources greatly improves monocular depth estimation. Our approach clearly outperforms competing methods across diverse datasets, setting a new state of the art for monocular depth estimation.

Today, embedded, mobile, and cyberphysical systems are ubiquitous and used in many applications, from industrial control systems, modern vehicles, to critical infrastructure. Current trends and initiatives, such as "Industrie 4.0" and Internet of Things (IoT), promise innovative business models and novel user experiences through strong connectivity and effective use of next generation of embedded devices. These systems generate, process, and exchange vast amounts of security-critical and privacy-sensitive data, which makes them attractive targets of attacks. Cyberattacks on IoT systems are very critical since they may cause physical damage and even threaten human lives. The complexity of these systems and the potential impact of cyberattacks bring upon new threats.

Deep-submicron technology is having a significant impact on permanent, intermittent, and transient classes of faults. This article discusses the main trends and challenges in circuit reliability, and explains evolving techniques for dealing with them.

Power forecasting of renewable energy power plants is a very active research field, as reliable information about the future power generation allow for a safe operation of the power grid and helps to minimize the operational costs of these energy sources. Deep Learning algorithms have shown to be very powerful in forecasting tasks, such as economic time series or speech recognition. Up to now, Deep Learning algorithms have only been applied sparsely for forecasting renewable energy power plants. By using different Deep Learning and Artificial Neural Network algorithms, such as Deep Belief Networks, AutoEncoder, and LSTM, we introduce these powerful algorithms in the field of renewable energy power forecasting. In our experiments, we used combinations of these algorithms to show their forecast strength compared to a standard MLP and a physical forecasting model in the forecasting the energy output of 21 solar power plants. Our results using Deep Learning algorithms show a superior forecasting performance compared to Artificial Neural Networks as well as other reference models such as physical models.

Abstract First-person view (FPV) drone racing is a televised sport in which professional competitors pilot high-speed aircraft through a 3D circuit. Each pilot sees the environment from the perspective of their drone by means of video streamed from an onboard camera. Reaching the level of professional pilots with an autonomous drone is challenging because the robot needs to fly at its physical limits while estimating its speed and location in the circuit exclusively from onboard sensors 1 . Here we introduce Swift, an autonomous system that can race physical vehicles at the level of the human world champions. The system combines deep reinforcement learning (RL) in simulation with data collected in the physical world. Swift competed against three human champions, including the world champions of two international leagues, in real-world head-to-head races. Swift won several races against each of the human champions and demonstrated the fastest recorded race time. This work represents a milestone for mobile robotics and machine intelligence 2 , which may inspire the deployment of hybrid learning-based solutions in other physical systems.

Many important applications trigger bulk bitwise operations, i.e., bitwise operations on large bit vectors. In fact, recent works design techniques that exploit fast bulk bitwise operations to accelerate databases (bitmap indices, BitWeaving) and web search (BitFunnel). Unfortunately, in existing architectures, the throughput of bulk bitwise operations is limited by the memory bandwidth available to the processing unit (e.g., CPU, GPU, FPGA, processing-in-memory).

Artificial Intelligence (AI) systems are embedded in a multifaceted environment characterized by intricate technical, legal, and organizational frameworks. To attain a comprehensive understanding of all AI-related risks, it is essential to evaluate both model-specific risks and those associated with the organizational and governance setups. We categorize these as "bottom-up risks" and "top-down risks," respectively. In this paper, we focus on the expansion and enhancement of a testing and auditing technology stack to identify and manage governance-related risks ("top-down"). These risks emerge from various dimensions, including internal development and decision-making processes, leadership structures, security setups, documentation practices, and more. For auditing governance related risk, we implement a traditional risk management framework and map it to the specifics of AI systems. Our end-to-end (from identification to monitoring) risk management kernel follows these implementation steps: - Identify - Collect - Assess - Comply - Monitor We demonstrate that scaling of such a risk auditing tool requires fundamental aspects. Those aspects include for instance a role-based approach, covering different roles in the development of complex AI systems. Ensuring compliance and secure record-keeping through audit-proof capabilities is also paramount. This ensures that the auditing technology can withstand scrutiny and maintain the integrity of records over time. Another critical aspect is the integrability of the auditing tool within existing risk management and governance infrastructures. This integration is essential to reduce the barriers for companies to comply with current regulatory requirements, such as the EU AI Act [European Parliament and the Council of the EU, 2024], and established standards like ISO 42001:2023. Ultimately, we demonstrate that this approach provides a robust technology stack for ensuring that AI systems are developed, utilized and supervised in a manner that is both compliant with regulatory standards and aligned with best practices in risk management and governance.

This paper describes a multi-core processor that integrates 48 cores, 4 DDR3 memory channels, and a voltage regulator controller in a 64 2D-mesh network-on-chip architecture. Located at each mesh node is a five-port virtual cut-through packet-switched router shared between two IA-32 cores. Core-to-core communication uses message passing while exploiting 384 KB of on-die shared memory. Fine grain power management takes advantage of 8 voltage and 28 frequency islands to allow independent DVFS of cores and mesh. At the nominal 1.1 V supply, the cores operate at 1 GHz while the 2D-mesh operates at 2 GHz. As performance and voltage scales, the processor dissipates between 25 W and 125 W. The processor is implemented in 45 nm Hi-K CMOS and has 1.3 billion transistors.

In speech emotion recognition, training and test data used for system development usually tend to fit each other perfectly, but further 'similar' data may be available. Transfer learning helps to exploit such similar data for training despite the inherent dissimilarities in order to boost a recogniser's performance. In this context, this paper presents a sparse auto encoder method for feature transfer learning for speech emotion recognition. In our proposed method, a common emotion-specific mapping rule is learnt from a small set of labelled data in a target domain. Then, newly reconstructed data are obtained by applying this rule on the emotion-specific data in a different domain. The experimental results evaluated on six standard databases show that our approach significantly improves the performance relative to learning each source domain independently.

Query optimization is one of the most challenging problems in database systems. Despite the progress made over the past decades, query optimizers remain extremely complex components that require a great deal of hand-tuning for specific workloads and datasets. Motivated by this shortcoming and inspired by recent advances in applying machine learning to data management challenges, we introduce Neo ( Neural Optimizer ), a novel learning-based query optimizer that relies on deep neural networks to generate query executions plans. Neo bootstraps its query optimization model from existing optimizers and continues to learn from incoming queries, building upon its successes and learning from its failures. Furthermore, Neo naturally adapts to underlying data patterns and is robust to estimation errors. Experimental results demonstrate that Neo, even when bootstrapped from a simple optimizer like PostgreSQL, can learn a model that offers similar performance to state-of-the-art commercial optimizers, and in some cases even surpass them.

The advent of Storage Class Memory (SCM) is driving a rethink of storage systems towards a single-level architecture where memory and storage are merged. In this context, several works have investigated how to design persistent trees in SCM as a fundamental building block for these novel systems. However, these trees are significantly slower than DRAM-based counterparts since trees are latency-sensitive and SCM exhibits higher latencies than DRAM. In this paper we propose a novel hybrid SCM-DRAM persistent and concurrent B-Tree, named Fingerprinting Persistent Tree (FPTree) that achieves similar performance to DRAM-based counterparts. In this novel design, leaf nodes are persisted in SCM while inner nodes are placed in DRAM and rebuilt upon recovery. The FPTree uses Fingerprinting, a technique that limits the expected number of in-leaf probed keys to one. In addition, we propose a hybrid concurrency scheme for the FPTree that is partially based on Hardware Transactional Memory. We conduct a thorough performance evaluation and show that the FPTree outperforms state-of-the-art persistent trees with different SCM latencies by up to a factor of 8.2. Moreover, we show that the FPTree scales very well on a machine with 88 logical cores. Finally, we integrate the evaluated trees in memcached and a prototype database. We show that the FPTree incurs an almost negligible performance overhead over using fully transient data structures, while significantly outperforming other persistent trees.

Quadrotors are agile. Unlike most other machines, they can traverse extremely complex environments at high speeds. To date, only expert human pilots have been able to fully exploit their capabilities. Autonomous operation with onboard sensing and computation has been limited to low speeds. State-of-the-art methods generally separate the navigation problem into subtasks: sensing, mapping, and planning. Although this approach has proven successful at low speeds, the separation it builds upon can be problematic for high-speed navigation in cluttered environments. The subtasks are executed sequentially, leading to increased processing latency and a compounding of errors through the pipeline. Here, we propose an end-to-end approach that can autonomously fly quadrotors through complex natural and human-made environments at high speeds with purely onboard sensing and computation. The key principle is to directly map noisy sensory observations to collision-free trajectories in a receding-horizon fashion. This direct mapping drastically reduces processing latency and increases robustness to noisy and incomplete perception. The sensorimotor mapping is performed by a convolutional network that is trained exclusively in simulation via privileged learning: imitating an expert with access to privileged information. By simulating realistic sensor noise, our approach achieves zero-shot transfer from simulation to challenging real-world environments that were never experienced during training: dense forests, snow-covered terrain, derailed trains, and collapsed buildings. Our work demonstrates that end-to-end policies trained in simulation enable high-speed autonomous flight through challenging environments, outperforming traditional obstacle avoidance pipelines.

Real-time recognition of dynamic hand gestures from video streams is a challenging task since (i) there is no indication when a gesture starts and ends in the video, (ii) performed gestures should only be recognized once, and (iii) the entire architecture should be designed considering the memory and power budget. In this work, we address these challenges by proposing a hierarchical structure enabling offline-working convolutional neural network (CNN) architectures to operate online efficiently by using sliding window approach. The proposed architecture consists of two models: (1) A detector which is a lightweight CNN architecture to detect gestures and (2) a classifier which is a deep CNN to classify the detected gestures. In order to evaluate the single-time activations of the detected gestures, we propose to use Levenshtein distance as an evaluation metric since it can measure misclassifications, multiple detections, and missing detections at the same time. We evaluate our architecture on two publicly available datasets - EgoGesture and NVIDIA Dynamic Hand Gesture Datasets - which require temporal detection and classification of the performed hand gestures. ResNeXt-101 model, which is used as a classifier, achieves the state-of-the-art offline classification accuracy of 94.04% and 83.82% for depth modality on EgoGesture and NVIDIA benchmarks, respectively. In real-time detection and classification, we obtain considerable early detections while achieving performances close to offline operation. The codes and pretrained models used in this work are publicly available <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> .

While 5G is being deployed and the economy and society begin to reap the associated benefits, the research and development community starts to focus on the next, 6th Generation (6G) of wireless communications. Although there are papers available in the literature on visions, requirements and technical enablers for 6G from various academic perspectives, there is a lack of joint industry and academic work towards 6G. In this paper a consolidated view on vision, values, use cases and key enabling technologies from leading industry stakeholders and academia is presented. The authors represent the mobile communications ecosystem with competences spanning hardware, link layer and networking aspects, as well as standardization and regulation. The second contribution of the paper is revisiting and analyzing the key concurrent initiatives on 6G. A third contribution of the paper is the identification and justification of six key 6G research challenges: (i) “connecting”, in the sense of empowering, exploiting and governing, intelligence; (ii) realizing a network of networks, i.e., leveraging on existing networks and investments, while reinventing roles and protocols where needed; (iii) delivering extreme experiences, when/where needed; (iv) (environmental, economic, social) sustainability to address the major challenges of current societies; (v) trustworthiness as an ingrained fundamental design principle; (vi) supporting cost-effective global service coverage. A fourth contribution is a comprehensive specification of a concrete first-set of industry and academia jointly defined use cases for 6G, e.g., massive twinning, cooperative robots, immersive telepresence, and others. Finally, the anticipated evolutions in the radio, network and management/orchestration domains are discussed.

In this article we introduce Simu5G, a new OMNeT++-based model library to simulate 5G networks. Simu5G allows users to simulate the data plane of 5G New Radio deployments, in an end-to-end perspective and including all protocol layers, making it a valuable tool for researchers and practitioners interested in the performance evaluation of 5G networks and services. We discuss the modelling of the protocol layers, network entities and functions, and validate our abstraction of the physical layer using 3GPP-based scenarios. Moreover, we show how Simu5G can be used to evaluate Multi-access Edge Computing (MEC) and Cellular Vehicle-to-everything (C-V2X) services offered through a 5G network.

Touch is a complex sensing modality owing to large number of receptors (mechano, thermal, pain) nonuniformly embedded in the soft skin all over the body. These receptors can gather and encode the large tactile data, allowing us to feel and perceive the real world. This efficient somatosensation far outperforms the touch-sensing capability of most of the state-of-the-art robots today and suggests the need for neural-like hardware for electronic skin (e-skin). This could be attained through either innovative schemes for developing distributed electronics or repurposing the neuromorphic circuits developed for other sensory modalities such as vision and audio. This Review highlights the hardware implementations of various computational building blocks for e-skin and the ways they can be integrated to potentially realize human skin-like or peripheral nervous system-like functionalities. The neural-like sensing and data processing are discussed along with various algorithms and hardware architectures. The integration of ultrathin neuromorphic chips for local computation and the printed electronics on soft substrate used for the development of e-skin over large areas are expected to advance robotic interaction as well as open new avenues for research in medical instrumentation, wearables, electronics, and neuroprosthetics.