Robot localization using a previously built map is essential for a variety of tasks including highly accurate navigation and mobile manipulation. A popular approach to robot localization is based on image-to-point cloud registration, which combines illumination-invariant LiDAR-based mapping with economical image-based localization. However, the recent works for image-to-point cloud registration either divide the registration into separate modules or project the point cloud to the depth image to register the RGB and depth images. In this paper, we present I2PNet, a novel end-to-end 2D-3D registration network. I2PNet directly registers the raw 3D point cloud with the 2D RGB image using differential modules with a unique target. The 2D-3D cost volume module for differential 2D-3D association is proposed to bridge feature extraction and pose regression. 2D-3D cost volume module implicitly constructs the soft point-to-pixel correspondence on the intrinsic-independent normalized plane of the pinhole camera model. Moreover, we introduce an outlier mask prediction module to filter the outliers in the 2D-3D association before pose regression. Furthermore, we propose the coarse-to-fine 2D-3D registration architecture to increase localization accuracy. We conduct extensive localization experiments on the KITTI Odometry and nuScenes datasets. The results demonstrate that I2PNet outperforms the state-of-the-art by a large margin. In addition, I2PNet has a higher efficiency than the previous works and can perform the localization in real-time. Moreover, we extend the application of I2PNet to the camera-LiDAR online calibration and demonstrate that I2PNet outperforms recent approaches on the online calibration task.
Cutting-edge connected vehicle (CV) technologies have drawn much attention in recent years. The real-time traffic data captured by a CV can be shared with other CVs and data centers so as to open new possibilities for solving diverse transportation problems. However, imagery captured by onboard cameras in a connected environment, are not sufficiently investigated, especially for safety and health-oriented visual perception. In this paper, a bidirectional process of image synthesis and decomposition (BPISD) approach is proposed, and thus a novel self-supervised multi-task learning framework, to simultaneously estimate depth map, atmospheric visibility, airlight, and PM2.5 mass concentration, in which depth map and visibility are considered highly associated with traffic safety, while airlight and PM2.5 mass concentration are directly correlated with human health. Both the training and testing phases of the proposed system solely require a single image as input. Due to the innovative training pipeline, the depth estimation network can manage various levels of visibility conditions and overcome inherent problems in current image-synthesis-based depth estimation, thereby generating high-quality depth maps even in low-visibility situations and further benefiting accurate estimations of visibility, airlight, and PM2.5 mass concentration. Extensive experiments on the synthesized data from the KITTI and real-world data collected in Beijing demonstrate that the proposed method can (1) achieve performance competitive in depth estimation as compared with state-of-the-art methods when taking clear images as input; (2) predict vivid depth map for images contaminated by various levels of haze; and (3) accurately estimate visibility, airlight, and PM2.5 mass concentrations. Beneficial applications can be developed based on the presented work to improve traffic safety, air quality, and public health.
The discussion on using zero padding (ZP) instead of a cyclic prefix (CP) for enhancing channel estimation and equalization performance is a recurring topic in waveform design for future wireless systems that high spectral efficiency and location awareness are the key factors. This is particularly true for orthogonal signals, such as orthogonal frequency-division multiplexing (OFDM). ZP-OFDM is appealing for joint communications and sensing (JCS) in 6G networks because it takes the advantage of both OFDM and pulse radar. In term of communication, ZP-OFDM compared to CP-OFDM, has higher power efficiency and lower bit error rate (BER). However, time synchronization is challenging in ZP-OFDM systems due to the lack of CP. In terms of sensing, ZP facilitates ranging methods, such as time-sum-of-arrival (TSOA). In this paper, we propose a moment-based timing offset (TO) estimator for multiple-input multiple-output (MIMO) ZP-OFDM system without the need for pilots. We then introduce the which significantly improves the estimation accuracy of the previous estimator. We show that the proposed method asymptotically reaches the maximum likelihood (ML) estimator. Simulation results show very high probability of lock-in for the proposed estimators under various practical scenarios.
In addition to its crucial impact on customer satisfaction, last-mile delivery (LMD) is notorious for being the most time-consuming and costly stage of the shipping process. Pressing environmental concerns combined with the recent surge of e-commerce sales have sparked renewed interest in automation and electrification of last-mile logistics. To address the hurdles faced by existing robotic couriers, this paper introduces a customer-centric and safety-conscious LMD system for small urban communities based on AI-assisted autonomous delivery robots. The presented framework enables end-to-end automation and optimization of the logistic process while catering for real-world imposed operational uncertainties, clients' preferred time schedules, and safety of pedestrians. To this end, the integrated optimization component is modeled as a robust variant of the Cumulative Capacitated Vehicle Routing Problem with Time Windows, where routes are constructed under uncertain travel times with an objective to minimize the total latency of deliveries (i.e., the overall waiting time of customers, which can negatively affect their satisfaction). We demonstrate the proposed LMD system's utility through real-world trials in a university campus with a single robotic courier. Implementation aspects as well as the findings and practical insights gained from the deployment are discussed in detail. Lastly, we round up the contributions with numerical simulations to investigate the scalability of the developed mathematical formulation with respect to the number of robotic vehicles and customers.
Meta-forecasting is a newly emerging field which combines meta-learning and time series forecasting. The goal of meta-forecasting is to train over a collection of source time series and generalize to new time series one-at-a-time. Previous approaches in meta-forecasting achieve competitive performance, but with the restriction of training a separate model for each sampling frequency. In this work, we investigate meta-forecasting over different sampling frequencies, and introduce a new model, the Continuous Frequency Adapter (CFA), specifically designed to learn frequency-invariant representations. We find that CFA greatly improves performance when generalizing to unseen frequencies, providing a first step towards forecasting over larger multi-frequency datasets.
The modeling of time-varying graph signals as stationary time-vertex stochastic processes permits the inference of missing signal values by efficiently employing the correlation patterns of the process across different graph nodes and time instants. In this study, we first propose an algorithm for computing graph autoregressive moving average (graph ARMA) processes based on learning the joint time-vertex power spectral density of the process from its incomplete realizations. Our solution relies on first roughly estimating the joint spectrum of the process from partially observed realizations and then refining this estimate by projecting it onto the spectrum manifold of the ARMA process. We then present a theoretical analysis of the sample complexity of learning graph ARMA processes. Experimental results show that the proposed approach achieves improvement in the time-vertex signal estimation performance in comparison with reference approaches in the literature.
Massive samples of event sequences data occur in various domains, including e-commerce, healthcare, and finance. There are two main challenges regarding inference of such data: computational and methodological. The amount of available data and the length of event sequences per client are typically large, thus it requires long-term modelling. Moreover, this data is often sparse and non-uniform, making classic approaches for time series processing inapplicable. Existing solutions include recurrent and transformer architectures in such cases. To allow continuous time, the authors introduce specific parametric intensity functions defined at each moment on top of existing models. Due to the parametric nature, these intensities represent only a limited class of event sequences. We propose the COTIC method based on a continuous convolution neural network suitable for non-uniform occurrence of events in time. In COTIC, dilations and multi-layer architecture efficiently handle dependencies between events. Furthermore, the model provides general intensity dynamics in continuous time - including self-excitement encountered in practice. The COTIC model outperforms existing approaches on majority of the considered datasets, producing embeddings for an event sequence that can be used to solve downstream tasks - e.g. predicting next event type and return time. The code of the proposed method can be found in the GitHub repository (https://github.com/VladislavZh/COTIC).
We introduce Cap3D, an automatic approach for generating descriptive text for 3D objects. This approach utilizes pretrained models from image captioning, image-text alignment, and LLM to consolidate captions from multiple views of a 3D asset, completely side-stepping the time-consuming and costly process of manual annotation. We apply Cap3D to the recently introduced large-scale 3D dataset, Objaverse, resulting in 660k 3D-text pairs. Our evaluation, conducted using 41k human annotations from the same dataset, demonstrates that Cap3D surpasses human-authored descriptions in terms of quality, cost, and speed. Through effective prompt engineering, Cap3D rivals human performance in generating geometric descriptions on 17k collected annotations from the ABO dataset. Finally, we finetune Text-to-3D models on Cap3D and human captions, and show Cap3D outperforms; and benchmark the SOTA including Point-E, Shape-E, and DreamFusion.
Nowadays, most 3D model quality assessment (3DQA) methods have been aimed at improving performance. However, little attention has been paid to the computational cost and inference time required for practical applications. Model-based 3DQA methods extract features directly from the 3D models, which are characterized by their high degree of complexity. As a result, many researchers are inclined towards utilizing projection-based 3DQA methods. Nevertheless, previous projection-based 3DQA methods directly extract features from multi-projections to ensure quality prediction accuracy, which calls for more resource consumption and inevitably leads to inefficiency. Thus in this paper, we address this challenge by proposing a no-reference (NR) projection-based \textit{\underline{G}rid \underline{M}ini-patch \underline{S}ampling \underline{3D} Model \underline{Q}uality \underline{A}ssessment (GMS-3DQA)} method. The projection images are rendered from six perpendicular viewpoints of the 3D model to cover sufficient quality information. To reduce redundancy and inference resources, we propose a multi-projection grid mini-patch sampling strategy (MP-GMS), which samples grid mini-patches from the multi-projections and forms the sampled grid mini-patches into one quality mini-patch map (QMM). The Swin-Transformer tiny backbone is then used to extract quality-aware features from the QMMs. The experimental results show that the proposed GMS-3DQA outperforms existing state-of-the-art NR-3DQA methods on the point cloud quality assessment databases. The efficiency analysis reveals that the proposed GMS-3DQA requires far less computational resources and inference time than other 3DQA competitors. The code will be available at https://github.com/zzc-1998/GMS-3DQA.
In this work, we introduce a ``score-based assessment'' framework for estimating the transferability of pre-trained speech models (PSMs) for fine-tuning target tasks. We leverage upon two representation theories, Bayesian likelihood estimation and optimal transport, to generate rank scores for the PSM candidates using the extracted representations. Our framework efficiently computes transferability scores without actual fine-tuning of candidate models or layers by making a temporal independent hypothesis. We evaluate some popular supervised speech models (e.g., Conformer RNN-Transducer) and self-supervised speech models (e.g., HuBERT) in cross-layer and cross-model settings using public data. Experimental results show a high Spearman's rank correlation and low $p$-value between our estimation framework and fine-tuning ground truth. Our proposed transferability framework requires less computational time and resources, making it a resource-saving and time-efficient approach for tuning speech foundation models.