Search

Robust grasping represents an essential task in robotics, necessitating tactile feedback and reactive grasping adjustments for robust grasping of objects. Previous research has extensively combined tactile sensing with grasping, primarily relying on rule-based approaches, frequently neglecting post-grasping difficulties such as external disruptions or inherent uncertainties of the object's physics and geometry. To address these limitations, this paper introduces an human-demonstration-based adaptive grasping policy base on tactile, which aims to achieve robust gripping while resisting disturbances to maintain grasp stability. Our trained model generalizes to daily objects with seven different sizes, shapes, and textures. Experimental results demonstrate that our method performs well in dynamic and force interaction tasks and exhibits excellent generalization ability.

Diffusion auction design for combinatorial settings is a long-standing challenge. One difficulty is that we cannot directly extend the solutions for simpler settings to combinatorial settings (like extending the Vickrey auction to VCG in the traditional settings). In this paper, we propose a different approach to leverage the diffusion auctions for single-item settings. We design a combinatorial diffusion auction framework which can use any desirable single-item diffusion auction to produce a combinatorial auction to satisfy incentive compatibility (IC), individual rationality (IR), and weak budget balance (WBB)., Comment: 8 pages

Soft object manipulation poses significant challenges for robots, requiring effective techniques for state representation and manipulation policy learning. State representation involves capturing the dynamic changes in the environment, while manipulation policy learning focuses on establishing the relationship between robot actions and state transformations to achieve specific goals. To address these challenges, this research paper introduces a novel approach: a dynamic heterogeneous graph-based model for learning goal-oriented soft object manipulation policies. The proposed model utilizes graphs as a unified representation for both states and policy learning. By leveraging the dynamic graph, we can extract crucial information regarding object dynamics and manipulation policies. Furthermore, the model facilitates the integration of demonstrations, enabling guided policy learning. To evaluate the efficacy of our approach, we designed a dough rolling task and conducted experiments using both a differentiable simulator and a real-world humanoid robot. Additionally, several ablation studies were performed to analyze the effect of our method, demonstrating its superiority in achieving human-like behavior., Comment: 7 pages, 5 figures Accepted by IEEE ROBIO 2024 conference

Causal language models acquire vast amount of knowledge from general text corpus during pretraining, but the efficiency of knowledge learning is known to be unsatisfactory, especially when learning from knowledge-dense and small-sized corpora. The deficiency can come from long-distance dependencies which are hard to capture by language models, and overfitting to co-occurrence patterns and distracting clues in the training text. To address these issues, the paper proposes a method to enhance knowledge learning during language model pretraining, by enhancing elusive but important clues in text discovered by the language model themselves. We found that larger language models pay more attention to non-obvious but important clues, which are often overlooked by smaller language models. Therefore, we can identify these clues by contrasting the attention weights of large and small language models. We use the identified clues as a guide to perform token-dropout data augmentation on the training text, and observed a significant boost in both small and large models' performance in fact memorization. This shows that the behavior contrast between more and less-performant language models contains important clues for knowledge learning, and it can be ``amplified" for a straight-forward improvement in knowledge learning efficiency., Comment: 7 pages and 17 figures

Multi-modal large language models (MLLMs) have shown impressive capabilities as a general-purpose interface for various visual and linguistic tasks. However, building a unified MLLM for multi-task learning in the medical field remains a thorny challenge. To mitigate the tug-of-war problem of multi-modal multi-task optimization in MLLMs, recent advances primarily focus on improving the LLM components, while neglecting the connector that bridges the gap between modalities. In this paper, we introduce Uni-Med, a novel medical generalist foundation model which consists of a universal visual feature extraction module, a connector mixture-of-experts (CMoE) module, and an LLM. Benefiting from the proposed CMoE that leverages a well-designed router with a mixture of projection experts at the connector, Uni-Med achieves efficient solution to the tug-of-war problem and can perform six different medical tasks including question answering, visual question answering, report generation, referring expression comprehension, referring expression generation and image classification. To the best of our knowledge, Uni-Med is the first effort to tackle multi-task interference at the connector in MLLMs. Extensive ablation experiments validate the effectiveness of introducing CMoE under any configuration, with up to an average 8% performance gains. We further provide interpretation analysis of the tug-of-war problem from the perspective of gradient optimization and parameter statistics. Compared to previous state-of-the-art medical MLLMs, Uni-Med achieves competitive or superior evaluation metrics on diverse tasks. Code and resources are available at https://github.com/tsinghua-msiip/Uni-Med.

Pretrained language models can encode a large amount of knowledge and utilize it for various reasoning tasks, yet they can still struggle to learn novel factual knowledge effectively from finetuning on limited textual demonstrations. In this work, we show that the reason for this deficiency is that language models are biased to learn word co-occurrence statistics instead of true factual associations. We identify the differences between two forms of knowledge representation in language models: knowledge in the form of co-occurrence statistics is encoded in the middle layers of the transformer model and does not generalize well to reasoning scenarios beyond simple question answering, while true factual associations are encoded in the lower layers and can be freely utilized in various reasoning tasks. Based on these observations, we propose two strategies to improve the learning of factual associations in language models. We show that training on text with implicit rather than explicit factual associations can force the model to learn factual associations instead of co-occurrence statistics, significantly improving the generalization of newly learned knowledge. We also propose a simple training method to actively forget the learned co-occurrence statistics, which unblocks and enhances the learning of factual associations when training on plain narrative text. On both synthetic and real-world corpora, the two proposed strategies improve the generalization of the knowledge learned during finetuning to reasoning scenarios such as indirect and multi-hop question answering.

In this paper, we briefly summarize the first competition on resource-limited infrared small target detection (namely, LimitIRSTD). This competition has two tracks, including weakly-supervised infrared small target detection (Track 1) and lightweight infrared small target detection (Track 2). 46 and 60 teams successfully registered and took part in Tracks 1 and Track 2, respectively. The top-performing methods and their results in each track are described with details. This competition inspires the community to explore the tough problems in the application of infrared small target detection, and ultimately promote the deployment of this technology under limited resource.

Robotic fish is one of the most promising directions of the new generation of underwater vehicles. Traditional biomimetic fish often mimic fish joints using tandem components like servos, which leads to increased volume, weight and control complexity. In this paper, a new double-joint robotic fish using a composite linkage was designed, where the propulsion mechanism transforms the single-degree-of-freedom rotation of the motor into a double-degree-of-freedom coupled motion, namely caudal peduncle translation and caudal fin rotation. Motion analysis of the propulsion mechanism demonstrates its ability to closely emulate the undulating movement observed in carangiform fish. Experimental results further validate the feasibility of the proposed propulsion mechanism. To improve propulsion efficiency, an analysis is conducted to explore the influence of swing angle amplitude and swing frequency on the swimming speed of the robotic fish. This examination establishes a practical foundation for future research on such robotic fish systems.

Using stacked emission line flux measurements of cool circumgalactic gas (CGM) in lower-mass galaxies ($10^{9.0} \le M_*/M_\odot \le 10^{10.2} $), we measure the dependence of the emission characteristics on orientation relative to the disk plane as a function of radius and compare to that we found previously for massive ($M_* > 10^{10.4} M_\odot$) early-type galaxies. Although the line ratios (the lower [N II]/H$\alpha$ and [O III]/H$\beta$) suggest an overall softer ionizing source than in the more massive galaxies, consistent with previous findings, we find the same ionization hardening signature (a higher [N II]/H$\alpha$ ratio in the inner polar region) along the polar direction at small radii that we found for the more massive galaxies. The line ratio in the inner polar bin is distinct from that measured for the inner planar bin with 99.99% confidence and with $>$ 99.9% confidence we conclude that it lies outside the star formation regime. The effective hardening of the ionization of the CGM along the polar axis, at small radii, could either indicate relic effects of AGN activity or shock ionization. In either case, this signature appears to be ubiquitous across the stellar mass range we are able to explore with our spectral stacking technique and currently available archival data., Comment: To appear in ApJ, 8 pages. arXiv admin note: text overlap with arXiv:2210.10043

Small object detection (SOD) has been a longstanding yet challenging task for decades, with numerous datasets and algorithms being developed. However, they mainly focus on either visible or thermal modality, while visible-thermal (RGBT) bimodality is rarely explored. Although some RGBT datasets have been developed recently, the insufficient quantity, limited category, misaligned images and large target size cannot provide an impartial benchmark to evaluate multi-category visible-thermal small object detection (RGBT SOD) algorithms. In this paper, we build the first large-scale benchmark with high diversity for RGBT SOD (namely RGBT-Tiny), including 115 paired sequences, 93K frames and 1.2M manual annotations. RGBT-Tiny contains abundant targets (7 categories) and high-diversity scenes (8 types that cover different illumination and density variations). Note that, over 81% of targets are smaller than 16x16, and we provide paired bounding box annotations with tracking ID to offer an extremely challenging benchmark with wide-range applications, such as RGBT fusion, detection and tracking. In addition, we propose a scale adaptive fitness (SAFit) measure that exhibits high robustness on both small and large targets. The proposed SAFit can provide reasonable performance evaluation and promote detection performance. Based on the proposed RGBT-Tiny dataset and SAFit measure, extensive evaluations have been conducted, including 23 recent state-of-the-art algorithms that cover four different types (i.e., visible generic detection, visible SOD, thermal SOD and RGBT object detection). Project is available at https://github.com/XinyiYing24/RGBT-Tiny.

Language models can learn sophisticated language understanding skills from fitting raw text. They also unselectively learn useless corpus statistics and biases, especially during finetuning on domain-specific corpora. In this paper, we propose a simple modification to causal language modeling called conditional finetuning, which performs language modeling conditioned on a context. We show that a context can "explain away" certain corpus statistics and make the model avoid learning them. In this fashion, conditional finetuning achieves selective learning from a corpus, learning knowledge useful for downstream tasks while avoiding learning useless corpus statistics like topic biases. This selective learning effect leads to less forgetting and better stability-plasticity tradeoff in domain finetuning, potentially benefitting lifelong learning with language models., Comment: To appear at the 41st International Conference on Machine Learning (ICML 2024)

We present TinyLLaVA Factory, an open-source modular codebase for small-scale large multimodal models (LMMs) with a focus on simplicity of code implementations, extensibility of new features, and reproducibility of training results. Following the design philosophy of the factory pattern in software engineering, TinyLLaVA Factory modularizes the entire system into interchangeable components, with each component integrating a suite of cutting-edge models and methods, meanwhile leaving room for extensions to more features. In addition to allowing users to customize their own LMMs, TinyLLaVA Factory provides popular training recipes to let users pretrain and finetune their models with less coding effort. Empirical experiments validate the effectiveness of our codebase. The goal of TinyLLaVA Factory is to assist researchers and practitioners in exploring the wide landscape of designing and training small-scale LMMs with affordable computational resources., Comment: Our codebase is made public at https://github.com/TinyLLaVA/TinyLLaVA_Factory with documentation available at https://tinyllava-factory.readthedocs.io/en/latest/

We present the results of direct-method metallicity measurements in the disk and outflow of the low-metallicity starburst galaxy NGC 1569. We use Keck Cosmic Web Imager observations to map the galaxy across 54$\arcsec$ (800 pc) along the major axis and 48$\arcsec$ (700 pc) along the minor axis with a spatial resolution of 1$\arcsec$ ($\sim$15 pc). We detect common strong emission lines ([\ion{O}{III}] $\lambda$5007, H$\beta$, [\ion{O}{II}] $\lambda$3727) and the fainter [\ion{O}{III}] $\lambda$4363 auroral line, which allows us to measure electron temperature ($T_e$) and metallicity. Theory suggests that outflows drive metals out of the disk driving observed trends between stellar mass and gas-phase metallicity. Our main result is that the metallicity in the outflow is similar to that of the disk, $Z_{\rm out} / Z_{\rm ISM} \approx 1$. This is consistent with previous absorption line studies in higher mass galaxies. Assumption of a mass-loading factor of $\dot{M}_{\rm out}/{\rm SFR}\sim3$ makes the metal-loading of NGC 1569 consistent with expectations derived from the mass-metallicity relationship. Our high spatial resolution metallicity maps reveal a region around a supermassive star cluster (SSC-B) with distinctly higher metallicity and higher electron density, compared to the disk. Given the known properties of SSC-B the higher metallicity and density of this region are likely the result of star formation-driven feedback acting on the local scale. Overall, our results are consistent with the picture in which metal-enriched winds pollute the circumgalactic medium surrounding galaxies, and thus connect the small-scale feedback processes to large-scale properties of galaxy halos., Comment: 15 pages, 11 figures, accepted by MNRAS

We present NewsBench, a novel evaluation framework to systematically assess the capabilities of Large Language Models (LLMs) for editorial capabilities in Chinese journalism. Our constructed benchmark dataset is focused on four facets of writing proficiency and six facets of safety adherence, and it comprises manually and carefully designed 1,267 test samples in the types of multiple choice questions and short answer questions for five editorial tasks in 24 news domains. To measure performances, we propose different GPT-4 based automatic evaluation protocols to assess LLM generations for short answer questions in terms of writing proficiency and safety adherence, and both are validated by the high correlations with human evaluations. Based on the systematic evaluation framework, we conduct a comprehensive analysis of ten popular LLMs which can handle Chinese. The experimental results highlight GPT-4 and ERNIE Bot as top performers, yet reveal a relative deficiency in journalistic safety adherence in creative writing tasks. Our findings also underscore the need for enhanced ethical guidance in machine-generated journalistic content, marking a step forward in aligning LLMs with journalistic standards and safety considerations., Comment: Long paper, ACL 2024 Main

Modern natural language generation systems with Large Language Models (LLMs) exhibit the capability to generate a plausible summary of multiple documents; however, it is uncertain if they truly possess the capability of information consolidation to generate summaries, especially on documents with opinionated information. We focus on meta-review generation, a form of sentiment summarisation for the scientific domain. To make scientific sentiment summarization more grounded, we hypothesize that human meta-reviewers follow a three-layer framework of sentiment consolidation to write meta-reviews. Based on the framework, we propose novel prompting methods for LLMs to generate meta-reviews and evaluation metrics to assess the quality of generated meta-reviews. Our framework is validated empirically as we find that prompting LLMs based on the framework -- compared with prompting them with simple instructions -- generates better meta-reviews., Comment: Long paper, ACL 2024 Main

We measure resolved (kiloparsec-scale) outflow properties in a sample of 10 starburst galaxies from the DUVET sample, using Keck/KCWI observations of H$\beta$ and [OIII]~$\lambda$5007. We measure $\sim450$ lines-of-sight that contain outflows, and use these to study scaling relationships of outflow velocity ($v_{\rm out}$), mass-loading factor ($\eta$; mass outflow rate per SFR) and mass flux ($\dot{\Sigma}_{\rm out}$; mass outflow rate per area) with co-located SFR surface density ($\Sigma_{\rm SFR}$) and stellar mass surface density ($\Sigma_{\ast}$). We find strong, positive correlations of $\dot{\Sigma}_{\rm out} \propto \Sigma_{\rm SFR}^{1.2}$ and $\dot{\Sigma}_{\rm out} \propto \Sigma_{\ast}^{1.7}$. We also find shallow correlations between $v_{\rm out}$ and both $\Sigma_{\rm SFR}$ and $\Sigma_{\ast}$. Our resolved observations do not suggest a threshold in outflows with $\Sigma_{\rm SFR}$, but rather we find that the local specific SFR ($\Sigma_{\rm SFR}/\Sigma_\ast$) is a better predictor of where outflows are detected. We find that outflows are very common above $\Sigma_{\rm SFR}/\Sigma_\ast\gtrsim 0.1$~Gyr$^{-1}$ and rare below this value. We argue that our results are consistent with a picture in which outflows are driven by supernovae, and require more significant injected energy in higher mass surface density environments to overcome local gravity. The correlations we present here provide a statistically robust, direct comparison for simulations and higher redshift results from JWST., Comment: 14 pages, 7 figures, plus 4 figures in appendix, submitted to MNRAS

Facilitating large-scale, cross-institutional collaboration in biomedical machine learning projects requires a trustworthy and resilient federated learning (FL) environment to ensure that sensitive information such as protected health information is kept confidential. In this work, we introduce APPFLx, a low-code FL framework that enables the easy setup, configuration, and running of FL experiments across organizational and administrative boundaries while providing secure end-to-end communication, privacy-preserving functionality, and identity management. APPFLx is completely agnostic to the underlying computational infrastructure of participating clients. We demonstrate the capability of APPFLx as an easy-to-use framework for accelerating biomedical studies across institutions and healthcare systems while maintaining the protection of private medical data in two case studies: (1) predicting participant age from electrocardiogram (ECG) waveforms, and (2) detecting COVID-19 disease from chest radiographs. These experiments were performed securely across heterogeneous compute resources, including a mixture of on-premise high-performance computing and cloud computing, and highlight the role of federated learning in improving model generalizability and performance when aggregating data from multiple healthcare systems. Finally, we demonstrate that APPFLx serves as a convenient and easy-to-use framework for accelerating biomedical studies across institutions and healthcare system while maintaining the protection of private medical data.

Generative models are a class of machine learning models that aim to learn the underlying probability distribution of data. Unlike discriminative models, generative models focus on capturing the data's inherent structure, allowing them to generate new samples that resemble the original data. To fully exploit the potential of modeling probability distributions using quantum physics, a quantum-inspired generative model known as the Born machines have shown great advancements in learning classical and quantum data over matrix product state(MPS) framework. The Born machines support tractable log-likelihood, autoregressive and mask sampling, and have shown outstanding performance in various unsupervised learning tasks. However, much of the current research has been centered on improving the expressive power of MPS, predominantly embedding each token directly by a corresponding tensor index. In this study, we generalize the embedding method into trainable quantum measurement operators that can be simultaneously honed with MPS. Our study indicated that combined with trainable embedding, Born machines can exhibit better performance and learn deeper correlations from the dataset., Comment: 5 pages, 4 figures

BACKGROUND & AIMS: Histological remission (HR) is advocated and considered as a new therapeutic target in ulcerative colitis (UC). Diagnosis of histologic remission currently relies on biopsy; during this process, patients are at risk for bleeding, infection, and post-biopsy fibrosis. In addition, histologic response scoring is complex and time-consuming, and there is heterogeneity among pathologists. Endocytoscopy (EC) is a novel ultra-high magnification endoscopic technique that can provide excellent in vivo assessment of glands. Based on the EC technique, we propose a neural network model that can assess histological disease activity in UC using EC images to address the above issues. The experiment results demonstrate that the proposed method can assist patients in precise treatment and prognostic assessment. METHODS: We construct a neural network model for UC evaluation. A total of 5105 images of 154 intestinal segments from 87 patients undergoing EC treatment at a center in China between March 2022 and March 2023 are scored according to the Geboes score. Subsequently, 103 intestinal segments are used as the training set, 16 intestinal segments are used as the validation set for neural network training, and the remaining 35 intestinal segments are used as the test set to measure the model performance together with the validation set. RESULTS: By treating HR as a negative category and histologic activity as a positive category, the proposed neural network model can achieve an accuracy of 0.9, a specificity of 0.95, a sensitivity of 0.75, and an area under the curve (AUC) of 0.81. CONCLUSION: We develop a specific neural network model that can distinguish histologic remission/activity in EC images of UC, which helps to accelerate clinical histological diagnosis. keywords: ulcerative colitis; Endocytoscopy; Geboes score; neural network.

We investigate the bilayer nickelates as a platform to realize the symmetric mass generation (SMG) insulator, a featureless Mott insulator that arises due to the Lieb-Schultz-Mattis (LSM) anomaly cancellation in bilayer spin-1/2 lattice systems. Through a single-orbital bilayer square lattice model involving intralayer hopping $t$ and interlayer superexchange interaction $J$, we demonstrate the emergence of high-temperature superconductivity (SC) upon doping the SMG insulator. The SC phase features $s$-wave interlayer spin-singlet pairing and exhibits a crossover between the BCS and BEC limits by tuning the $J/t$ ratio. We estimate the SC transition temperature $T_c$ from both the weak and strong coupling limits at the mean-field level. Our findings offer insights into the experimentally observed decrease in $T_c$ with pressure and the strange metal behavior above $T_c$. Additionally, we propose that both Ni $3d_{z^2}$ and $3d_{x^2-y^2}$ orbitals can exhibit superconductivity in La$_3$Ni$_2$O$_7$ under pressure, but their $T_c$ should vary in opposite ways under doping. This characteristic difference suggests a potential experimental pathway to identify which electronic orbital plays the principal role in the formation of superconductivity in this system., Comment: 11 pages, 5 figures, 2 tables

We present a method to characterize star-formation driven outflows from edge-on galaxies and apply this method to the metal-poor starburst galaxy, Mrk 1486. Our method uses the distribution of emission line flux (from H$\beta$ and [OIII] 5007) to identify the location of the outflow and measure the extent above the disk, the opening angle, and the transverse kinematics. We show that this simple technique recovers a similar distribution of the outflow without requiring complex modelling of line-splitting or multi-Gaussian components, and is therefore applicable to lower spectral resolution data. In Mrk 1486 we observe an asymmetric outflow in both the location of the peak flux and total flux from each lobe. We estimate an opening angle of $17-37^{\circ}$ depending on the method and assumptions adopted. Within the minor axis outflows, we estimate a total mass outflow rate of $\sim2.5$ M$_{\odot}$ yr$^{-1}$, which corresponds to a mass loading factor of $\eta=0.7$. We observe a non-negligible amount of flux from ionized gas outflowing along the edge of the disk (perpendicular to the biconical components), with a mass outflow rate $\sim0.9$ M$_{\odot}$ yr$^{-1}$. Our results are intended to demonstrate a method that can be applied to high-throughput, low spectral resolution observations, such as narrow band filters or low spectral resolution IFS that may be more able to recover the faint emission from outflows., Comment: 12 Pages, 6 Figures

During the process of robot-assisted ultrasound(US) puncture, it is important to estimate the location of the puncture from the 2D US images. To this end, the calibration of the US image becomes an important issue. In this paper, we proposed a depth camera-based US calibration method, where an easy-to-deploy device is designed for the calibration. With this device, the coordinates of the puncture needle tip are collected respectively in US image and in the depth camera, upon which a correspondence matrix is built for calibration. Finally, a number of experiments are conducted to validate the effectiveness of our calibration method.

Tactile sensing is of great importance during human hand usage such as object exploration, grasping and manipulation. Different types of tactile sensors have been designed during the past decades, which are mainly focused on either the fingertips for grasping or the upper-body for human-robot interaction. In this paper, a novel soft tactile sensor has been designed to mimic the functionality of human palm that can estimate the contact state of different objects. The tactile palm mainly consists of three parts including an electrode array, a soft cover skin and the conductive sponge. The design principle are described in details, with a number of experiments showcasing the effectiveness of the proposed design.

As medical ultrasound is becoming a prevailing examination approach nowadays, robotic ultrasound systems can facilitate the scanning process and prevent professional sonographers from repetitive and tedious work. Despite the recent progress, it is still a challenge to enable robots to autonomously accomplish the ultrasound examination, which is largely due to the lack of a proper task representation method, and also an adaptation approach to generalize learned skills across different patients. To solve these problems, we propose the latent task representation and the robotic skills adaptation for autonomous ultrasound in this paper. During the offline stage, the multimodal ultrasound skills are merged and encapsulated into a low-dimensional probability model through a fully self-supervised framework, which takes clinically demonstrated ultrasound images, probe orientations, and contact forces into account. During the online stage, the probability model will select and evaluate the optimal prediction. For unstable singularities, the adaptive optimizer fine-tunes them to near and stable predictions in high-confidence regions. Experimental results show that the proposed approach can generate complex ultrasound strategies for diverse populations and achieve significantly better quantitative results than our previous method.

For the SLAM system in robotics and autonomous driving, the accuracy of front-end odometry and back-end loop-closure detection determine the whole intelligent system performance. But the LiDAR-SLAM could be disturbed by current scene moving objects, resulting in drift errors and even loop-closure failure. Thus, the ability to detect and segment moving objects is essential for high-precision positioning and building a consistent map. In this paper, we address the problem of moving object segmentation from 3D LiDAR scans to improve the odometry and loop-closure accuracy of SLAM. We propose a novel 3D Sequential Moving-Object-Segmentation (3D-SeqMOS) method that can accurately segment the scene into moving and static objects, such as moving and static cars. Different from the existing projected-image method, we process the raw 3D point cloud and build a 3D convolution neural network for MOS task. In addition, to make full use of the spatio-temporal information of point cloud, we propose a point cloud residual mechanism using the spatial features of current scan and the temporal features of previous residual scans. Besides, we build a complete SLAM framework to verify the effectiveness and accuracy of 3D-SeqMOS. Experiments on SemanticKITTI dataset show that our proposed 3D-SeqMOS method can effectively detect moving objects and improve the accuracy of LiDAR odometry and loop-closure detection. The test results show our 3D-SeqMOS outperforms the state-of-the-art method by 12.4%. We extend the proposed method to the SemanticKITTI: Moving Object Segmentation competition and achieve the 2nd in the leaderboard, showing its effectiveness.

The Hot Universe Baryon Surveyor (HUBS) is a proposed space-based X-ray telescope for detecting X-ray emissions from the hot gas content in our universe. With its unprecedented spatially-resolved high-resolution spectroscopy and large field of view, the HUBS mission will be uniquely qualified to measure the physical and chemical properties of the hot gas in the interstellar medium, the circumgalactic medium, the intergalactic medium, and the intracluster medium. These measurements will be valuable for two key scientific goals of HUBS, namely to unravel the AGN and stellar feedback physics that governs the formation and evolution of galaxies, and to probe the baryon budget and multi-phase states from galactic to cosmological scales. In addition to these two goals, the HUBS mission will also help us solve some problems in the fields of galaxy clusters, AGNs, diffuse X-ray backgrounds, supernova remnants, and compact objects. This paper discusses the perspective of advancing these fields using the HUBS telescope., Comment: 52 pages, 22 figures. Accepted for publication in Science China: Physics, Mechanics and Astronomy

High-velocity atomic clouds in the Galactic center have attracted significant attention due to their enigmatic formation process, which is potentially linked to the starburst or supermassive black hole activities in the region. Further, the discovery of high-velocity molecular clouds (HVMCs) presents a greater puzzle, because they are much denser and more massive. If the HVMCs were accelerated by the strong activities in the Galactic center, they are expected to be destroyed before they reach such a high velocity. To shed light on this phenomenon, we perform three-dimensional numerical simulations to investigate the origin and hydrodynamic evolution of HVMCs during a starburst in the Galactic center. We find that the presence of a magnetic field provides effective protection and acceleration to molecular clouds (MCs) within the galactic winds. Consequently, the MCs can attain latitudes of approximately 1 kpc with velocities around 200 km/s, consistent with the observed characteristics of HVMCs. The consistency of our findings across a wide parameter space supports the conclusion that HVMCs can indeed withstand the starburst environment in the Galactic center, providing valuable insights into their survival mechanisms., Comment: 18 pages, 16 figures, submitted to MNRAS

The prominence of embodied Artificial Intelligence (AI), which empowers robots to navigate, perceive, and engage within virtual environments, has attracted significant attention, owing to the remarkable advancements in computer vision and large language models. Privacy emerges as a pivotal concern within the realm of embodied AI, as the robot accesses substantial personal information. However, the issue of privacy leakage in embodied AI tasks, particularly in relation to reinforcement learning algorithms, has not received adequate consideration in research. This paper aims to address this gap by proposing an attack on the value-based algorithm and the gradient-based algorithm, utilizing gradient inversion to reconstruct states, actions, and supervision signals. The choice of using gradients for the attack is motivated by the fact that commonly employed federated learning techniques solely utilize gradients computed based on private user data to optimize models, without storing or transmitting the data to public servers. Nevertheless, these gradients contain sufficient information to potentially expose private data. To validate our approach, we conduct experiments on the AI2THOR simulator and evaluate our algorithm on active perception, a prevalent task in embodied AI. The experimental results demonstrate the effectiveness of our method in successfully reconstructing all information from the data across 120 room layouts., Comment: 7 pages, 4 figures, 2 tables

The NLI4CT task aims to entail hypotheses based on Clinical Trial Reports (CTRs) and retrieve the corresponding evidence supporting the justification. This task poses a significant challenge, as verifying hypotheses in the NLI4CT task requires the integration of multiple pieces of evidence from one or two CTR(s) and the application of diverse levels of reasoning, including textual and numerical. To address these problems, we present a multi-granularity system for CTR-based textual entailment and evidence retrieval in this paper. Specifically, we construct a Multi-granularity Inference Network (MGNet) that exploits sentence-level and token-level encoding to handle both textual entailment and evidence retrieval tasks. Moreover, we enhance the numerical inference capability of the system by leveraging a T5-based model, SciFive, which is pre-trained on the medical corpus. Model ensembling and a joint inference method are further utilized in the system to increase the stability and consistency of inference. The system achieves f1-scores of 0.856 and 0.853 on textual entailment and evidence retrieval tasks, resulting in the best performance on both subtasks. The experimental results corroborate the effectiveness of our proposed method. Our code is publicly available at https://github.com/THUMLP/NLI4CT., Comment: Accepted by SemEval2023