In future 6G Mobile Edge Computing (MEC), autopilot systems require the capability of processing multimodal data with strong interdependencies. However, traditional heuristic algorithms are inadequate for real-time scheduling due to their requirement for multiple iterations to derive the optimal scheme. We propose a novel TSNet-SAC based on Transformer, that utilizes heuristic algorithms solely to guide the training of TSNet. Additionally, a Sliding Augment Component (SAC) is introduced to enhance the robustness and resolve algorithm defects. Furthermore, the Extender component is designed to handle multi-scale training data and provide network scalability, enabling TSNet to adapt to different access scenarios. Simulation demonstrates that TSNet-SAC outperforms existing networks in accuracy and robustness, achieving superior scheduling-making latency compared to heuristic algorithms.

The current era of natural language processing (NLP) has been defined by the prominence of pre-trained language models since the advent of BERT. A feature of BERT and models with similar architecture is the objective of masked language modeling, in which part of the input is intentionally masked and the model is trained to predict this piece of masked information. Data augmentation is a data-driven technique widely used in machine learning, including research areas like computer vision and natural language processing, to improve model performance by artificially augmenting the training data set by designated techniques. Masked language models (MLM), an essential training feature of BERT, have introduced a novel approach to perform effective pre-training on Transformer based models in natural language processing tasks. Recent studies have utilized masked language model to generate artificially augmented data for NLP downstream tasks. The experimental results show that Mask based data augmentation method provides a simple but efficient approach to improve the model performance. In this paper, we explore and discuss the broader utilization of these data augmentation methods based on MLM.

We propose a general strategy for autonomous guidance and insertion of a needle into a retinal blood vessel. The main challenges underpinning this task are the accurate placement of the needle-tip on the target vein and a careful needle insertion maneuver to avoid double-puncturing the vein, while dealing with challenging kinematic constraints and depth-estimation uncertainty. Following how surgeons perform this task purely based on visual feedback, we develop a system which relies solely on \emph{monocular} visual cues by combining data-driven kinematic and contact estimation, visual-servoing, and model-based optimal control. By relying on both known kinematic models, as well as deep-learning based perception modules, the system can localize the surgical needle tip and detect needle-tissue interactions and venipuncture events. The outputs from these perception modules are then combined with a motion planning framework that uses visual-servoing and optimal control to cannulate the target vein, while respecting kinematic constraints that consider the safety of the procedure. We demonstrate that we can reliably and consistently perform needle insertion in the domain of retinal surgery, specifically in performing retinal vein cannulation. Using cadaveric pig eyes, we demonstrate that our system can navigate to target veins within 22$\mu m$ XY accuracy and perform the entire procedure in less than 35 seconds on average, and all 24 trials performed on 4 pig eyes were successful. Preliminary comparison study against a human operator show that our system is consistently more accurate and safer, especially during safety-critical needle-tissue interactions. To the best of the authors' knowledge, this work accomplishes a first demonstration of autonomous retinal vein cannulation at a clinically-relevant setting using animal tissues.

POI tagging aims to annotate a point of interest (POI) with some informative tags, which facilitates many services related to POIs, including search, recommendation, and so on. Most of the existing solutions neglect the significance of POI images and seldom fuse the textual and visual features of POIs, resulting in suboptimal tagging performance. In this paper, we propose a novel Multi-Modal Model for POI Tagging, namely M3PT, which achieves enhanced POI tagging through fusing the target POI's textual and visual features, and the precise matching between the multi-modal representations. Specifically, we first devise a domain-adaptive image encoder (DIE) to obtain the image embeddings aligned to their gold tags' semantics. Then, in M3PT's text-image fusion module (TIF), the textual and visual representations are fully fused into the POIs' content embeddings for the subsequent matching. In addition, we adopt a contrastive learning strategy to further bridge the gap between the representations of different modalities. To evaluate the tagging models' performance, we have constructed two high-quality POI tagging datasets from the real-world business scenario of Ali Fliggy. Upon the datasets, we conducted the extensive experiments to demonstrate our model's advantage over the baselines of uni-modality and multi-modality, and verify the effectiveness of important components in M3PT, including DIE, TIF and the contrastive learning strategy.

The superior performance of Deep Neural Networks (DNNs) has led to their application in various aspects of human life. Safety-critical applications are no exception and impose rigorous reliability requirements on DNNs. Quantized Neural Networks (QNNs) have emerged to tackle the complexity of DNN accelerators, however, they are more prone to reliability issues. In this paper, a recent analytical resilience assessment method is adapted for QNNs to identify critical neurons based on a Neuron Vulnerability Factor (NVF). Thereafter, a novel method for splitting the critical neurons is proposed that enables the design of a Lightweight Correction Unit (LCU) in the accelerator without redesigning its computational part. The method is validated by experiments on different QNNs and datasets. The results demonstrate that the proposed method for correcting the faults has a twice smaller overhead than a selective Triple Modular Redundancy (TMR) while achieving a similar level of fault resiliency.

Orthogonality regularization has been developed to prevent deep CNNs from training instability and feature redundancy. Among existing proposals, kernel orthogonality regularization enforces orthogonality by minimizing the residual between the Gram matrix formed by convolutional filters and the orthogonality matrix. We propose a novel measure for achieving better orthogonality among filters, which disentangles diagonal and correlation information from the residual. The model equipped with the measure under the principle of imposing strict orthogonality between filters surpasses previous regularization methods in near-orthogonality. Moreover, we observe the benefits of improved strict filter orthogonality in relatively shallow models, but as model depth increases, the performance gains in models employing strict kernel orthogonality decrease sharply. Furthermore, based on the observation of the potential conflict between strict kernel orthogonality and growing model capacity, we propose a relaxation theory on kernel orthogonality regularization. The relaxed kernel orthogonality achieves enhanced performance on models with increased capacity, shedding light on the burden of strict kernel orthogonality on deep model performance. We conduct extensive experiments with our kernel orthogonality regularization toolkit on ResNet and WideResNet in CIFAR-10 and CIFAR-100. We observe state-of-the-art gains in model performance from the toolkit, which includes both strict orthogonality and relaxed orthogonality regularization, and obtain more robust models with expressive features. These experiments demonstrate the efficacy of our toolkit and subtly provide insights into the often overlooked challenges posed by strict orthogonality, addressing the burden of strict orthogonality on capacity-rich models.

Natural Language Inference (NLI) has been a cornerstone task in evaluating language models' inferential reasoning capabilities. However, the standard three-way classification scheme used in NLI has well-known shortcomings in evaluating models' ability to capture the nuances of natural human reasoning. In this paper, we argue that the operationalization of the neutral label in current NLI datasets has low validity, is interpreted inconsistently, and that at least one important sense of neutrality is often ignored. We uncover the detrimental impact of these shortcomings, which in some cases leads to annotation datasets that actually decrease performance on downstream tasks. We compare approaches of handling annotator disagreement and identify flaws in a recent NLI dataset that designs an annotator study based on a problematic operationalization. Our findings highlight the need for a more refined evaluation framework for NLI, and we hope to spark further discussion and action in the NLP community.

Can robots mold soft plastic materials by shaping depth images? The short answer is no: current day robots can't. In this article, we address the problem of shaping plastic material with an anthropomorphic arm/hand robot, which observes the material with a fixed depth camera. Robots capable of molding could assist humans in many tasks, such as cooking, scooping or gardening. Yet, the problem is complex, due to its high-dimensionality at both perception and control levels. To address it, we design three alternative data-based methods for predicting the effect of robot actions on the material. Then, the robot can plan the sequence of actions and their positions, to mold the material into a desired shape. To make the prediction problem tractable, we rely on two original ideas. First, we prove that under reasonable assumptions, the shaping problem can be mapped from point cloud to depth image space, with many benefits (simpler processing, no need for registration, lower computation time and memory requirements). Second, we design a novel, simple metric for quickly measuring the distance between two depth images. The metric is based on the inherent point cloud representation of depth images, which enables direct and consistent comparison of image pairs through a non-uniform scaling approach, and therefore opens promising perspectives for designing \textit{depth image -- based} robot controllers. We assess our approach in a series of unprecedented experiments, where a robotic arm/hand molds flour from initial to final shapes, either with its own dataset, or by transfer learning from a human dataset. We conclude the article by discussing the limitations of our framework and those of current day hardware, which make human-like robot molding a challenging open research problem.

We consider the problem of subset selection where one is given multiple rankings of items and the goal is to select the highest ``quality'' subset. Score functions from the multiwinner voting literature have been used to aggregate rankings into quality scores for subsets. We study this setting of subset selection problems when, in addition, rankings may contain systemic or unconscious biases toward a group of items. For a general model of input rankings and biases, we show that requiring the selected subset to satisfy group fairness constraints can improve the quality of the selection with respect to unbiased rankings. Importantly, we show that for fairness constraints to be effective, different multiwinner score functions may require a drastically different number of rankings: While for some functions, fairness constraints need an exponential number of rankings to recover a close-to-optimal solution, for others, this dependency is only polynomial. This result relies on a novel notion of ``smoothness'' of submodular functions in this setting that quantifies how well a function can ``correctly'' assess the quality of items in the presence of bias. The results in this paper can be used to guide the choice of multiwinner score functions for the subset selection setting considered here; we additionally provide a tool to empirically enable this.

Model distillation has emerged as a prominent technique to improve neural search models. To date, distillation taken an offline approach, wherein a new neural model is trained to predict relevance scores between arbitrary queries and documents. In this paper, we explore a departure from this offline distillation strategy by investigating whether a model for a specific query can be effectively distilled from neural re-ranking results (i.e., distilling in an online setting). Indeed, we find that a lexical model distilled online can reasonably replicate the re-ranking of a neural model. More importantly, these models can be used as queries that execute efficiently on indexes. This second retrieval stage can enrich the pool of documents for re-ranking by identifying documents that were missed in the first retrieval stage. Empirically, we show that this approach performs favourably when compared with established pseudo relevance feedback techniques, dense retrieval methods, and sparse-dense ensemble "hybrid" approaches.