Arm DevSummit was one of several virtual tech conferences and launch events taking place over the past month week and it included everything from a fireside chat with the CEOs of Arm and Nvidia about the proposed acquisition to new technology announcements and the deep dive training tracks one would expect. One of the most interesting announcements was around a new platform of cores for SoCs specifically targeting autonomous machines. While there is a growing focus on autonomous vehicles, few companies look at the broader opportunity for autonomous control across a wide variety of platforms that include autonomous systems for materials handling and production for everything from consumer products to agriculture. Arm announced a complete platform that can span a wide range of applications that meet the safety critical requirements of autonomous systems. As Arm expanded its technology offerings in mobile, it began releasing a complete platform of compute cores and related technologies together as a single platform.

Many real-world systems, such as moving planets, can be considered as multi-agent dynamic systems, where objects interact with each other and co-evolve along with the time. Such dynamics is usually difficult to capture, and understanding and predicting the dynamics based on observed trajectories of objects become a critical research problem in many domains. Most existing algorithms, however, assume the observations are regularly sampled and all the objects can be fully observed at each sampling time, which is impractical for many applications. In this paper, we propose to learn system dynamics from irregularly-sampled partial observations with underlying graph structure for the first time. To tackle the above challenge, we present LG-ODE, a latent ordinary differential equation generative model for modeling multi-agent dynamic system with known graph structure. It can simultaneously learn the embedding of high dimensional trajectories and infer continuous latent system dynamics. Our model employs a novel encoder parameterized by a graph neural network that can infer initial states in an unsupervised way from irregularly-sampled partial observations of structural objects and utilizes neural ODE to infer arbitrarily complex continuous-time latent dynamics. Experiments on motion capture, spring system, and charged particle datasets demonstrate the effectiveness of our approach.

The widespread interest in autonomous driving technology in recent years [2] has motivated extensive research in multiagent navigation in driving domains. One of the most challenging driving domains [3] is the uncontrolled intersection, i.e., a street intersection that features no traffic signs or signals. Within this domain, we focus on scenarios in which agents do not communicate explicitly or implicitly through e.g., turn signals. This model setup gives rise to challenging multi-vehicle encounters that mimic real-world situations (arising due to human distraction, violation of traffic rules or special emergencies) that result in fatal accidents [3]. The frequency and severity of such situations has motivated vivid research interest in uncontrolled intersections [4, 5, 6]. In the absence of explicit traffic signs, signals, rules or explicit communication among agents, avoiding collisions at intersections relies on the ability of agents to predict the dynamics of interaction amongst themselves. One prevalent way to model multiagent dynamics is via trajectory prediction. However, multistep multiagent trajectory prediction is NPhard [7], whereas the sample complexity of existing learning algorithms effectively prohibits the extraction of practical models. Our key insight is that the geometric structure of the intersection and the incentive of agents to move efficiently and avoid collisions with each other (rationality) compress the space of possible multiagent trajectories, effectively simplifying inference.

Artificial Intelligence (AI) or Machine Learning (ML) systems have been widely adopted as value propositions by companies in all industries in order to create or extend the services and products they offer. However, developing AI/ML systems has presented several engineering problems that are different from those that arise in, non-AI/ML software development. This study aims to investigate how software engineering (SE) has been applied in the development of AI/ML systems and identify challenges and practices that are applicable and determine whether they meet the needs of professionals. Also, we assessed whether these SE practices apply to different contexts, and in which areas they may be applicable. We conducted a systematic review of literature from 1990 to 2019 to (i) understand and summarize the current state of the art in this field and (ii) analyze its limitations and open challenges that will drive future research. Our results show these systems are developed on a lab context or a large company and followed a research-driven development process. The main challenges faced by professionals are in areas of testing, AI software quality, and data management. The contribution types of most of the proposed SE practices are guidelines, lessons learned, and tools.

Target-driven visual navigation is a challenging problem that requires a robot to find the goal using only visual inputs. Many researchers have demonstrated promising results using deep reinforcement learning (deep RL) on various robotic platforms, but typical end-to-end learning is known for its poor extrapolation capability to new scenarios. Therefore, learning a navigation policy for a new robot with a new sensor configuration or a new target still remains a challenging problem. In this paper, we introduce a learning algorithm that enables rapid adaptation to new sensor configurations or target objects with a few shots. We design a policy architecture with latent features between perception and inference networks and quickly adapt the perception network via meta-learning while freezing the inference network. Our experiments show that our algorithm adapts the learned navigation policy with only three shots for unseen situations with different sensor configurations or different target colors. We also analyze the proposed algorithm by investigating various hyperparameters.

You can find more information about formatting under the section "Submission guidelines" https://www.springer.com/journal/146. For inquiries and to submit your abstract and manuscript, please contact: aisocietyncstate@gmail.com

Multiplayer games have long been used as testbeds in artificial intelligence research, aptly referred to as the Drosophila of artificial intelligence. Traditionally, researchers have focused on using well-known games to build strong agents. This progress, however, can be better informed by characterizing games and their topological landscape. Tackling this latter question can facilitate understanding of agents and help determine what game an agent should target next as part of its training. Here, we show how network measures applied to response graphs of large-scale games enable the creation of a landscape of games, quantifying relationships between games of varying sizes and characteristics. We illustrate our findings in domains ranging from canonical games to complex empirical games capturing the performance of trained agents pitted against one another. Our results culminate in a demonstration leveraging this information to generate new and interesting games, including mixtures of empirical games synthesized from real world games. Multiplayer games can be used as testbeds for the development of learning algorithms for artificial intelligence. Omidshafiei et al. show how to characterize and compare such games using a graph-based approach, generating new games that could potentially be interesting for training in a curriculum.

Key Takeaways: # Intelligent software agents must use common sense in order to reason. Common-sense knowledge is required before intelligent software agents can anticipate how people and the physical world react. Deep learning models do not currently understand what they produce, and have no common-sense knowledge. The Commonsense Transformers (COMET) project attempts to train models with information about the world in ways similar to how a human would acquire such knowledge. The COMET project and other similar efforts are still in the research phase.

Reward shaping is an effective technique for incorporating domain knowledge into reinforcement learning (RL). Existing approaches such as potential-based reward shaping normally make full use of a given shaping reward function. However, since the transformation of human knowledge into numeric reward values is often imperfect due to reasons such as human cognitive bias, completely utilizing the shaping reward function may fail to improve the performance of RL algorithms. In this paper, we consider the problem of adaptively utilizing a given shaping reward function. We formulate the utilization of shaping rewards as a bi-level optimization problem, where the lower level is to optimize policy using the shaping rewards and the upper level is to optimize a parameterized shaping weight function for true reward maximization. We formally derive the gradient of the expected true reward with respect to the shaping weight function parameters and accordingly propose three learning algorithms based on different assumptions. Experiments in sparse-reward cartpole and MuJoCo environments show that our algorithms can fully exploit beneficial shaping rewards, and meanwhile ignore unbeneficial shaping rewards or even transform them into beneficial ones.

Many important robotics problems are partially observable in the sense that a single visual or force-feedback measurement is insufficient to reconstruct the state. Standard approaches involve learning a policy over beliefs or observation-action histories. However, both of these have drawbacks; it is expensive to track the belief online, and it is hard to learn policies directly over histories. We propose a method for policy learning under partial observability called the Belief-Grounded Network (BGN) in which an auxiliary belief-reconstruction loss incentivizes a neural network to concisely summarize its input history. Since the resulting policy is a function of the history rather than the belief, it can be executed easily at runtime. We compare BGN against several baselines on classic benchmark tasks as well as three novel robotic touch-sensing tasks. BGN outperforms all other tested methods and its learned policies work well when transferred onto a physical robot.