English has long been assumed the $\textit{lingua franca}$ of scientific research, and this notion is reflected in the natural language processing (NLP) research involving scientific document representation. In this position piece, we quantitatively show that the literature is largely multilingual and argue that current models and benchmarks should reflect this linguistic diversity. We provide evidence that text-based models fail to create meaningful representations for non-English papers and highlight the negative user-facing impacts of using English-only models non-discriminately across a multilingual domain. We end with suggestions for the NLP community on how to improve performance on non-English documents.
Data annotation in autonomous vehicles is a critical step in the development of Deep Neural Network (DNN) based models or the performance evaluation of the perception system. This often takes the form of adding 3D bounding boxes on time-sequential and registered series of point-sets captured from active sensors like Light Detection and Ranging (LiDAR) and Radio Detection and Ranging (RADAR). When annotating multiple active sensors, there is a need to motion compensate and translate the points to a consistent coordinate frame and timestamp respectively. However, highly dynamic objects pose a unique challenge, as they can appear at different timestamps in each sensor's data. Without knowing the speed of the objects, their position appears to be different in different sensor outputs. Thus, even after motion compensation, highly dynamic objects are not matched from multiple sensors in the same frame, and human annotators struggle to add unique bounding boxes that capture all objects. This article focuses on addressing this challenge, primarily within the context of Scania collected datasets. The proposed solution takes a track of an annotated object as input and uses the Moving Horizon Estimation (MHE) to robustly estimate its speed. The estimated speed profile is utilized to correct the position of the annotated box and add boxes to object clusters missed by the original annotation.
This paper presents two algorithms for multi-agent dynamic coverage in spatiotemporal environments, where the coverage algorithms are informed by the method of data assimilation. In particular, we show that by explicitly modeling the environment using a Gaussian Process (GP) model, and considering the sensing capabilities and the dynamics of a team of robots, we can design an estimation algorithm and multi-agent coverage controller that explores and estimates the state of the spatiotemporal environment. The uncertainty of the estimate is quantified using clarity, an information-theoretic metric, where higher clarity corresponds to lower uncertainty. By exploiting the relationship between GPs and Stochastic Differential Equations (SDEs) we quantify the increase in clarity of the estimated state at any position due to a measurement taken from any other position. We use this relationship to design two new coverage controllers, both of which scale well with the number of agents exploring the domain, assuming the robots can share the map of the clarity over the spatial domain via communication. We demonstrate the algorithms through a realistic simulation of a team of robots collecting wind data over a region in Austria.
We study the problem of mobile manipulation using legged robots equipped with an arm, namely legged loco-manipulation. The robot legs, while usually utilized for mobility, offer an opportunity to amplify the manipulation capabilities by conducting whole-body control. That is, the robot can control the legs and the arm at the same time to extend its workspace. We propose a framework that can conduct the whole-body control autonomously with visual observations. Our approach, namely Visual Whole-Body Control(VBC), is composed of a low-level policy using all degrees of freedom to track the body velocities along with the end-effector position, and a high-level policy proposing the velocities and end-effector position based on visual inputs. We train both levels of policies in simulation and perform Sim2Real transfer for real robot deployment. We perform extensive experiments and show significant improvements over baselines in picking up diverse objects in different configurations (heights, locations, orientations) and environments.
Failure Mode, Effects and Criticality Analysis (FMECA) is one of the safety analysis methods recommended by most of the international standards. The classical FMECA is made in a form of a table filled in either manually or by using safety analysis tools. In both cases, the design engineers have to choose the trade-offs between safety and other development constraints. In the case of complex cyber-physical systems (CPS) with thousands of specified constraints, this may lead to severe problems and significantly impact the overall criticality of CPS. In this paper, we propose to adopt optimization techniques to automate the decision making process conducted after FMECA of CPS. We describe a multi-agent based optimization method which extends classical FMECA for offering optimal solutions in terms of criticality and development constraints of CPS.
In this project, we examine how freelancers changed their strategic positioning on an online work platform following the launch of ChatGPT in November 2022 - a major advance in AI technologies. We document that post-ChatGPT, freelancers bid on fewer jobs and reposition themselves by differentiating their distribution of bids (i.e., job applications) relative to their prior behavior. We disentangle heterogeneity in strategic responses by exploring how exposure to changes in demand or supply shape incumbent repositioning. We find that the launch of ChatGPT was associated with a short-term decrease in labor demand and an increase in labor supply, though these changes vary across work domains. In response to decreases in labor demand, workers changed their horizontal positioning and withdrew from the platform. In response to increases in labor supply, workers were less likely to decrease bidding or reposition horizontally but shifted their vertical position by targeting lower-value jobs. We further show that repositioning is less likely for high-skill freelancers who face greater adjustment costs. This research contributes to our understanding of how and why workers respond to technological change in the context of recent advances in AI technologies.
In transformer architectures, position encoding primarily provides a sense of sequence for input tokens. While the original transformer paper's method has shown satisfactory results in general language processing tasks, there have been new proposals, such as Rotary Position Embedding (RoPE), for further improvement. This paper presents geotokens, input components for transformers, each linked to a specific geological location. Unlike typical language sequences, for these tokens, the order is not as vital as the geographical coordinates themselves. To represent the relative position in this context and to keep a balance between the real world distance and the distance in the embedding space, we design a position encoding approach drawing from the RoPE structure but tailored for spherical coordinates.
Pre-training strategies based on self-supervised learning (SSL) have proven to be effective pretext tasks for many downstream tasks in computer vision. Due to the significant disparity between medical and natural images, the application of typical SSL is not straightforward in medical imaging. Additionally, those pretext tasks often lack context, which is critical for computer-aided clinical decision support. In this paper, we developed a longitudinal masked auto-encoder (MAE) based on the well-known Transformer-based MAE. In particular, we explored the importance of time-aware position embedding as well as disease progression-aware masking. Taking into account the time between examinations instead of just scheduling them offers the benefit of capturing temporal changes and trends. The masking strategy, for its part, evolves during follow-up to better capture pathological changes, ensuring a more accurate assessment of disease progression. Using OPHDIAT, a large follow-up screening dataset targeting diabetic retinopathy (DR), we evaluated the pre-trained weights on a longitudinal task, which is to predict the severity label of the next visit within 3 years based on the past time series examinations. Our results demonstrated the relevancy of both time-aware position embedding and masking strategies based on disease progression knowledge. Compared to popular baseline models and standard longitudinal Transformers, these simple yet effective extensions significantly enhance the predictive ability of deep classification models.
In this paper, we address the challenge of making ViT models more robust to unseen affine transformations. Such robustness becomes useful in various recognition tasks such as face recognition when image alignment failures occur. We propose a novel method called KP-RPE, which leverages key points (e.g.~facial landmarks) to make ViT more resilient to scale, translation, and pose variations. We begin with the observation that Relative Position Encoding (RPE) is a good way to bring affine transform generalization to ViTs. RPE, however, can only inject the model with prior knowledge that nearby pixels are more important than far pixels. Keypoint RPE (KP-RPE) is an extension of this principle, where the significance of pixels is not solely dictated by their proximity but also by their relative positions to specific keypoints within the image. By anchoring the significance of pixels around keypoints, the model can more effectively retain spatial relationships, even when those relationships are disrupted by affine transformations. We show the merit of KP-RPE in face and gait recognition. The experimental results demonstrate the effectiveness in improving face recognition performance from low-quality images, particularly where alignment is prone to failure. Code and pre-trained models are available.
Point cloud processing methods leverage local and global point features %at the feature level to cater to downstream tasks, yet they often overlook the task-level context inherent in point clouds during the encoding stage. We argue that integrating task-level information into the encoding stage significantly enhances performance. To that end, we propose SMTransformer which incorporates task-level information into a vector-based transformer by utilizing a soft mask generated from task-level queries and keys to learn the attention weights. Additionally, to facilitate effective communication between features from the encoding and decoding layers in high-level tasks such as segmentation, we introduce a skip-attention-based up-sampling block. This block dynamically fuses features from various resolution points across the encoding and decoding layers. To mitigate the increase in network parameters and training time resulting from the complexity of the aforementioned blocks, we propose a novel shared position encoding strategy. This strategy allows various transformer blocks to share the same position information over the same resolution points, thereby reducing network parameters and training time without compromising accuracy.Experimental comparisons with existing methods on multiple datasets demonstrate the efficacy of SMTransformer and skip-attention-based up-sampling for point cloud processing tasks, including semantic segmentation and classification. In particular, we achieve state-of-the-art semantic segmentation results of 73.4% mIoU on S3DIS Area 5 and 62.4% mIoU on SWAN dataset
3D cell tracking in a living organism has a crucial role in live cell image analysis. Cell tracking in C. elegans has two difficulties. First, cell migration in a consecutive frame is large since they move their head during scanning. Second, cell detection is often inconsistent in consecutive frames due to touching cells and low-contrast images, and these inconsistent detections affect the tracking performance worse. In this paper, we propose a cell tracking method to address these issues, which has two main contributions. First, we introduce cell position heatmap-based non-rigid alignment with test-time fine-tuning, which can warp the detected points to near the positions at the next frame. Second, we propose a pairwise detection method, which uses the information of detection results at the previous frame for detecting cells at the current frame. The experimental results demonstrate the effectiveness of each module, and the proposed method achieved the best performance in comparison.
As an essential task in autonomous driving (AD), motion prediction aims to predict the future states of surround objects for navigation. One natural solution is to estimate the position of other agents in a step-by-step manner where each predicted time-step is conditioned on both observed time-steps and previously predicted time-steps, i.e., autoregressive prediction. Pioneering works like SocialLSTM and MFP design their decoders based on this intuition. However, almost all state-of-the-art works assume that all predicted time-steps are independent conditioned on observed time-steps, where they use a single linear layer to generate positions of all time-steps simultaneously. They dominate most motion prediction leaderboards due to the simplicity of training MLPs compared to autoregressive networks. In this paper, we introduce the GPT style next token prediction into motion forecasting. In this way, the input and output could be represented in a unified space and thus the autoregressive prediction becomes more feasible. However, different from language data which is composed of homogeneous units -words, the elements in the driving scene could have complex spatial-temporal and semantic relations. To this end, we propose to adopt three factorized attention modules with different neighbors for information aggregation and different position encoding styles to capture their relations, e.g., encoding the transformation between coordinate systems for spatial relativity while adopting RoPE for temporal relativity. Empirically, by equipping with the aforementioned tailored designs, the proposed method achieves state-of-the-art performance in the Waymo Open Motion and Waymo Interaction datasets. Notably, AMP outperforms other recent autoregressive motion prediction methods: MotionLM and StateTransformer, which demonstrates the effectiveness of the proposed designs.
Rotary Position Embedding (RoPE) performs remarkably on language models, especially for length extrapolation of Transformers. However, the impacts of RoPE on computer vision domains have been underexplored, even though RoPE appears capable of enhancing Vision Transformer (ViT) performance in a way similar to the language domain. This study provides a comprehensive analysis of RoPE when applied to ViTs, utilizing practical implementations of RoPE for 2D vision data. The analysis reveals that RoPE demonstrates impressive extrapolation performance, i.e., maintaining precision while increasing image resolution at inference. It eventually leads to performance improvement for ImageNet-1k, COCO detection, and ADE-20k segmentation. We believe this study provides thorough guidelines to apply RoPE into ViT, promising improved backbone performance with minimal extra computational overhead. Our code and pre-trained models are available at https://github.com/naver-ai/rope-vit
This paper addresses the problem of simultaneous estimation of the position, linear velocity and orientation of a rigid body using single bearing measurements. We introduce a Riccati observer-based estimator that fuses measurements from a 3-axis accelerometer, a 3-axis gyroscope, a single body-frame vector observation (e.g., magnetometer), and a single bearing-to-landmark measurement to obtain the full vehicle's state (position, velocity, orientation). The proposed observer guarantees global exponential convergence under some persistency of excitation (PE) condition on the vehicle's motion. Simulation results are presented to show the effectiveness of the proposed approach.
We formulate a novel planar motion planning problem for a Dubins-Laser system that consists of a Dubins vehicle with an attached controllable laser. The vehicle moves with unit speed and the laser, having a finite range, can rotate in a clockwise or anti-clockwise direction with a bounded angular rate. From an arbitrary initial position and orientation, the objective is to steer the system so that a given static target is within the range of the laser and the laser is oriented at it in minimum time. We characterize multiple properties of the optimal trajectory and establish that the optimal trajectory for the Dubins-laser system is one out of a total of 16 candidates. Finally, we provide numerical insights that illustrate the properties characterized in this work.
We propose an efficient symbolic control synthesis algorithm for equivariant continuous-time dynamical systems to satisfy reach-avoid specifications. The algorithm exploits dynamical symmetries to construct lean abstractions to avoid redundant computations during synthesis. Our proposed algorithm adds another layer of abstraction over the common grid-based discrete abstraction before solving the synthesis problem. It combines each set of grid cells that are at a similar relative position from the targets and nearby obstacles, defined by the symmetries, into a single abstract state. It uses this layer of abstraction to guide the order by which actions are explored during synthesis over the grid-based abstraction. We demonstrate the potential of our algorithm by synthesizing a reach-avoid controller for a 3-dimensional ship model with translation and rotation symmetries in the special Euclidean group SE(2).
We present the POLAR Traverse Dataset: a dataset of high-fidelity stereo pair images of lunar-like terrain under polar lighting conditions designed to simulate a straight-line traverse. Images from individual traverses with different camera heights and pitches were recorded at 1 m intervals by moving a suspended stereo bar across a test bed filled with regolith simulant and shaped to mimic lunar south polar terrain. Ground truth geometry and camera position information was also recorded. This dataset is intended for developing and testing software algorithms that rely on stereo or monocular camera images, such as visual odometry, for use in the lunar polar environment, as well as to provide insight into the expected lighting conditions in lunar polar regions.
Despite recent efforts to develop large language models with robust long-context capabilities, the lack of long-context benchmarks means that relatively little is known about their performance. To alleviate this gap, in this paper, we propose \textbf{Counting-Stars}, a multi-evidence, position-aware, and scalable benchmark designed to evaluate the multi-evidence retrieval capabilities of long-context LLMs. \textbf{Counting-Stars} comprises two counting-based multiple pieces of evidence retrieval sub-tasks: searching and reasoning. Using Counting-Stars, we conduct experiments to evaluate several long-context LLMs, including GPT-4 Turbo, Gemini 1.5 Pro, Claude3 Opus, GLM-4, and Moonshot-v1. Extensive experimental results demonstrate that Gemini 1.5 Pro achieves the best overall results, while GPT-4 Turbo exhibits the most stable performance across various tasks. Furthermore, our analysis of these LLMs, which have been extended to handle long-context scenarios, indicates that significant room for improvement remains as the length of the input context and the complexity of the tasks increase.
We consider both $N$-player and mean-field games of optimal portfolio liquidation in which the players are not allowed to change the direction of trading. Players with an initially short position of stocks are only allowed to buy while players with an initially long position are only allowed to sell the stock. Under suitable conditions on the model parameters we show that the games are equivalent to games of timing where the players need to determine the optimal times of market entry and exit. We identify the equilibrium entry and exit times and prove that equilibrium mean-trading rates can be characterized in terms of the solutions to a highly non-linear higher-order integral equation with endogenous terminal condition. We prove the existence of a unique solution to the integral equation from which we obtain the existence of a unique equilibrium both in the mean-field and the $N$-player game.
The line-of-sight (LOS) requirement of free-space optical (FSO) systems can be relaxed by employing optical intelligent reflecting surfaces (IRSs). In this paper, we model the impact of the IRS-induced delay dispersion and derive the channel impulse response (CIR) of IRS-assisted FSO links. The proposed model takes into account the characteristics of the incident and reflected beams' wavefronts, the position of transmitter and receiver, the size of the IRS, and the incident beamwidth on the IRS. Our simulation results reveal that a maximum effective delay spread of 0.7 ns is expected for a square IRS with area 1 $\mathrm{m}^2$, which induces inter-symbol interference for bit rates larger than 10 Gbps. We show that the IRS-induced delay dispersion can be mitigated via equalization at the receiver.