We develop data-driven models to predict when a robot should feed during social dining scenarios. Being able to eat independently with friends and family is considered one of the most memorable and important activities for people with mobility limitations. While existing robotic systems for feeding people with mobility limitations focus on solitary dining, commensality, the act of eating together, is often the practice of choice. Sharing meals with others introduces the problem of socially appropriate bite timing for a robot, i.e. the appropriate timing for the robot to feed without disrupting the social dynamics of a shared meal. Our key insight is that bite timing strategies that take into account the delicate balance of social cues can lead to seamless interactions during robot-assisted feeding in a social dining scenario. We approach this problem by collecting a Human-Human Commensality Dataset (HHCD) containing 30 groups of three people eating together. We use this dataset to analyze human-human commensality behaviors and develop bite timing prediction models in social dining scenarios. We also transfer these models to human-robot commensality scenarios. Our user studies show that prediction improves when our algorithm uses multimodal social signaling cues between diners to model bite timing. The HHCD dataset, videos of user studies, and code are available at https://emprise.cs.cornell.edu/hrcom/

AutoDRIVE is envisioned to be a comprehensive research platform for scaled autonomous vehicles. This work is a stepping-stone towards the greater goal of realizing such a research platform. Particularly, this work proposes a pseudo-realistic simulator for scaled autonomous vehicles, which is targeted towards simplicity, modularity and flexibility. The AutoDRIVE Simulator not only mimics realistic system dynamics but also simulates a comprehensive sensor suite and realistic actuator response. The simulator also features a communication bridge in order to interface externally developed autonomous driving software stack, which allows users to design and develop their algorithms conveniently and have them tested on our simulator. Presently, the bridge is compatible with Robot Operating System (ROS) and can be interfaced directly with the Python and C++ scripts developed as a part of this project. The bridge supports local as well as distributed computing.

This course is all about data and how it is critical to the success of your applied machine learning model. Understand the critical elements of data in the learning, training and operation phases Understand biases and sources of data Implement techniques to improve the generality of your model Explain the consequences of overfitting and identify mitigation measures Implement appropriate test and validation measures. Demonstrate how the accuracy of your model can be improved with thoughtful feature engineering. Explore the impact of the algorithm parameters on model strength To be successful in this course, you should have at least beginner-level background in Python programming (e.g., be able to read and code trace existing code, be comfortable with conditionals, loops, variables, lists, dictionaries and arrays). You should have a basic understanding of linear algebra (vector notation) and statistics (probability distributions and mean/median/mode).

On November 14th 2022, MeetKai, the LA-based AI and metaverse company held an extraordinary meeting with Kura Kura Bali, within the 17th G20 Heads of State and Government Summit, taking place in Bali this year. In an agreement between MeetKai Inc. and the CEO of PT Bali Turtle Island Development, Aaron Tham, the tech startup agreed to establish an R&D product development center in Kura Kura Bali focusing on software development and data science, as well as metaverse, AI and VR services. As hinted in their August New York metaverse launch, where a hyper-realistic anamorphic animation of Bali boasted "A Better Metaverse is Coming" in the middle of Times Square, MeetKai intends to "bring an inclusive metaverse to life", where wonders of the real world become accessible to everyone, everywhere. This Monday, MeetKai Co-Founders, serial entrepreneur Weili Dai, and CEO James Kaplan, agreed to contribute e-learning technology and training to support education in Bali. The company intends to support UID Bali Campus with connection and capacity building, subject expert matter, and synergy for events taking place in Kura Kura Bali; and most importantly, to contribute to their training data scientist programme by assuring cloud platform capacity and other machine learning skills to develop and strengthen the human resources skills in data science.

Large Language Models (LLMs) have been transformative. They are pre-trained foundational models that are self-supervised and can be adapted with fine tuning to a wide range of natural language tasks, each of which previously would have required a separate network model. This is one step closer to the extraordinary versatility of human language. GPT-3 and more recently LaMDA can carry on dialogs with humans on many topics after minimal priming with a few examples. However, there has been a wide range of reactions and debate on whether these LLMs understand what they are saying or exhibit signs of intelligence. This high variance is exhibited in three interviews with LLMs reaching wildly different conclusions. A new possibility was uncovered that could explain this divergence. What appears to be intelligence in LLMs may in fact be a mirror that reflects the intelligence of the interviewer, a remarkable twist that could be considered a Reverse Turing Test. If so, then by studying interviews we may be learning more about the intelligence and beliefs of the interviewer than the intelligence of the LLMs. As LLMs become more capable they may transform the way we interact with machines and how they interact with each other. Increasingly, LLMs are being coupled with sensorimotor devices. LLMs can talk the talk, but can they walk the walk? A road map for achieving artificial general autonomy is outlined with seven major improvements inspired by brain systems. LLMs could be used to uncover new insights into brain function by downloading brain data during natural behaviors.

Deep learning has achieved great success in learning features from massive remote sensing images (RSIs). To better understand the connection between feature learning paradigms (e.g., unsupervised feature learning (USFL), supervised feature learning (SFL), and self-supervised feature learning (SSFL)), this paper analyzes and compares them from the perspective of feature learning signals, and gives a unified feature learning framework. Under this unified framework, we analyze the advantages of SSFL over the other two learning paradigms in RSIs understanding tasks and give a comprehensive review of the existing SSFL work in RS, including the pre-training dataset, self-supervised feature learning signals, and the evaluation methods. We further analyze the effect of SSFL signals and pre-training data on the learned features to provide insights for improving the RSI feature learning. Finally, we briefly discuss some open problems and possible research directions.

Powder X-ray diffraction analysis is a critical component of materials characterization methodologies. Discerning characteristic Bragg intensity peaks and assigning them to known crystalline phases is the first qualitative step of evaluating diffraction spectra. Subsequent to phase identification, Rietveld refinement may be employed to extract the abundance of quantitative, material-specific parameters hidden within powder data. These characterization procedures are yet time-consuming and inhibit efficiency in materials science workflows. The ever-increasing popularity and propulsion of data science techniques has provided an obvious solution on the course towards materials analysis automation. Deep learning has become a prime focus for predicting crystallographic parameters and features from X-ray spectra. However, the infeasibility of curating large, well-labelled experimental datasets means that one must resort to a large number of theoretic simulations for powder data augmentation to effectively train deep models. Herein, we are interested in conventional supervised learning algorithms in lieu of deep learning for multi-label crystalline phase identification and quantitative phase analysis for a biomedical application. First, models were trained using very limited experimental data. Further, we incorporated simulated XRD data to assess model generalizability as well as the efficacy of simulation-based training for predictive analysis in a real-world X-ray diffraction application.

Humans tame the complexity of mathematical reasoning by developing hierarchies of abstractions. With proper abstractions, solutions to hard problems can be expressed concisely, thus making them more likely to be found. In this paper, we propose Learning Mathematical Abstractions (LEMMA): an algorithm that implements this idea for reinforcement learning agents in mathematical domains. LEMMA augments Expert Iteration with an abstraction step, where solutions found so far are revisited and rewritten in terms of new higher-level actions, which then become available to solve new problems. We evaluate LEMMA on two mathematical reasoning tasks--equation solving and fraction simplification--in a step-by-step fashion. In these two domains, LEMMA improves the ability of an existing agent, both solving more problems and generalizing more effectively to harder problems than those seen during training.

Building models that can be rapidly adapted to novel tasks using only a handful of annotated examples is an open challenge for multimodal machine learning research. We introduce Flamingo, a family of Visual Language Models (VLM) with this ability. We propose key architectural innovations to: (i) bridge powerful pretrained vision-only and language-only models, (ii) handle sequences of arbitrarily interleaved visual and textual data, and (iii) seamlessly ingest images or videos as inputs. Thanks to their flexibility, Flamingo models can be trained on large-scale multimodal web corpora containing arbitrarily interleaved text and images, which is key to endow them with in-context few-shot learning capabilities. We perform a thorough evaluation of our models, exploring and measuring their ability to rapidly adapt to a variety of image and video tasks. These include open-ended tasks such as visual question-answering, where the model is prompted with a question which it has to answer; captioning tasks, which evaluate the ability to describe a scene or an event; and close-ended tasks such as multiple-choice visual question-answering. For tasks lying anywhere on this spectrum, a single Flamingo model can achieve a new state of the art with few-shot learning, simply by prompting the model with task-specific examples. On numerous benchmarks, Flamingo outperforms models fine-tuned on thousands of times more task-specific data.

Continuous-time event sequences, i.e., sequences consisting of continuous time stamps and associated event types ("marks"), are an important type of sequential data with many applications, e.g., in clinical medicine or user behavior modeling. Since these data are typically modeled autoregressively (e.g., using neural Hawkes processes or their classical counterparts), it is natural to ask questions about future scenarios such as "what kind of event will occur next" or "will an event of type $A$ occur before one of type $B$". Unfortunately, some of these queries are notoriously hard to address since current methods are limited to naive simulation, which can be highly inefficient. This paper introduces a new typology of query types and a framework for addressing them using importance sampling. Example queries include predicting the $n^\text{th}$ event type in a sequence and the hitting time distribution of one or more event types. We also leverage these findings further to be applicable for estimating general "$A$ before $B$" type of queries. We prove theoretically that our estimation method is effectively always better than naive simulation and show empirically based on three real-world datasets that it is on average 1,000 times more efficient than existing approaches.