Search

Multiple Perspective attack Investigation (MPI) is a technique to partition application dependencies based on high-level semantics. It facilitates provenance analysis by generating succinct causal graphs. It involves an annotation process that identifies variables and data structures corresponding to the partitions and the communication channels between them. Though the amount of annotation is small, this process requires a detailed understanding of the source code. In this work, autoMPI, we extend the capability of MPI by automating the identifying annotation requirements. We leverage a hybrid analysis approach, performing a differential analysis based on crafted inputs. Static analysis is conducted to identify the annotation sites within the application code afterward automatically. Our evaluation shows the proposed approach can significantly facilitate the annotation process. It correctly identifies all required annotation sites within an average 16 seconds analysis time for the majority of analyzed programs with average precision and recall 72.5% and 100%, respectively.

Sushi, von Willebrand factor type A, EGF pentraxin domain-containing 1 (SVEP1), an extracellular matrix protein, is a human coronary artery disease locus that promotes atherosclerosis. We previously demonstrated that SVEP1 induces vascular smooth muscle cell (VSMC) proliferation and an inflammatory phenotype in the arterial wall to enhance the development of atherosclerotic plaque. The only receptor known to interact with SVEP1 is integrin α9β1, a cell surface receptor that is expressed by VSMCs and myeloid lineage-derived monocytes and macrophages. Our previous in vitro studies suggested that integrin α9β1 was necessary for SVEP1-induced VSMC proliferation and inflammation; however, the underlying mechanisms mediated by integrin α9β1 in these cell types during the development of atherosclerosis remain poorly understood.Here, using cell-specific gene targeting, we investigated the effects of the integrin α9β1 receptor on VSMCs and myeloid cells in mouse models of atherosclerosis. Interestingly, we found that depleting integrin α9β1 in either VSMCs or myeloid cells did not affect the formation or complexity of atherosclerotic plaque in vessels after either 8 or 16 weeks of high fat diet feeding.Our results indicate that integrin α9β1 in these two cell types does not mediate the in vivo effect of SVEP1 in the development of atherosclerosis. Instead, our results suggest either the presence of other potential receptor(s) or alternative integrin α9β1-expressing cell types responsible for SVEP1 induced signaling in the development of atherosclerosis.

Referring video object segmentation (RVOS) aims at segmenting an object in a video following human instruction. Current state-of-the-art methods fall into an offline pattern, in which each clip independently interacts with text embedding for cross-modal understanding. They usually present that the offline pattern is necessary for RVOS, yet model limited temporal association within each clip. In this work, we break up the previous offline belief and propose a simple yet effective online model using explicit query propagation, named OnlineRefer. Specifically, our approach leverages target cues that gather semantic information and position prior to improve the accuracy and ease of referring predictions for the current frame. Furthermore, we generalize our online model into a semi-online framework to be compatible with video-based backbones. To show the effectiveness of our method, we evaluate it on four benchmarks, \ie, Refer-Youtube-VOS, Refer-DAVIS17, A2D-Sentences, and JHMDB-Sentences. Without bells and whistles, our OnlineRefer with a Swin-L backbone achieves 63.5 J&F and 64.8 J&F on Refer-Youtube-VOS and Refer-DAVIS17, outperforming all other offline methods., Accepted by ICCV2023. The code is at https://github.com/wudongming97/OnlineRefer

In multi-timescale multi-agent reinforcement learning (MARL), agents interact across different timescales. In general, policies for time-dependent behaviors, such as those induced by multiple timescales, are non-stationary. Learning non-stationary policies is challenging and typically requires sophisticated or inefficient algorithms. Motivated by the prevalence of this control problem in real-world complex systems, we introduce a simple framework for learning non-stationary policies for multi-timescale MARL. Our approach uses available information about agent timescales to define a periodic time encoding. In detail, we theoretically demonstrate that the effects of non-stationarity introduced by multiple timescales can be learned by a periodic multi-agent policy. To learn such policies, we propose a policy gradient algorithm that parameterizes the actor and critic with phase-functioned neural networks, which provide an inductive bias for periodicity. The framework's ability to effectively learn multi-timescale policies is validated on a gridworld and building energy management environment., Accepted at IEEE CDC'23. 7 pages, 6 figures

Yan Zhang,1,2 Yang Liu,1,2 Xiangyu Zhang,1,2 V Wee Yong,3 Mengzhou Xue1,2 1Department of Cerebrovascular Diseases, The Second Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan, People’s Republic of China; 2Academy of Medical Science, Zhengzhou University, Zhengzhou, Henan, People’s Republic of China; 3Department of Clinical Neurosciences, University of Calgary, Calgary, Alberta, CanadaCorrespondence: V Wee Yong, Hotchkiss Brain Institute and Department of Clinical Neurosciences, University of Calgary, Calgary, Alberta, Canada, Email vyong@ucalgary.ca Mengzhou Xue, Department of Cerebrovascular Diseases, The Second Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan, 450001, People’s Republic of China, Email xuemengzhou@zzu.edu.cnPurpose: Intracerebral hemorrhage (ICH) is a fatal disease without effective treatment. The damage of the blood–brain barrier (BBB) is a key cause of brain edema and herniation after ICH. Omarigliptin (also known as MK3102) is a potent antidiabetic that inhibits dipeptidyl peptidase (DPP4); the latter has the ability to bind and degrade matrix metalloproteinases (MMPs). The present study aims to investigate the protective effects of omarigliptin against the destruction of BBB following ICH in mice.Methods and Materials: Collagenase VII was used to induce ICH in C57BL/6 mice. MK3102 (7 mg/kg/day) was administered after ICH. The modified neurological severity scores (mNSS) were carried out to assess neurological functions. Nissl staining was applied to evaluate neuronal loss. Brain water content, Evans blue extravasation, Western blots, immunohistochemistry and immunofluorescence were used to study the protective effects of BBB with MK3102 at 3 days after ICH.Results: MK3102 reduced DPP4 expression and decreased hematoma formation and neurobehavioral deficits of ICH mice. This was correspondent with lowered activation of microglia/macrophages and infiltration of neutrophils after ICH. Importantly, MK3102 protected the integrity of the BBB after ICH, associated with decreased expression of MMP-9, and preservation of the tight junction proteins ZO-1 and Occludin on endothelial cells through putative degradation of MMP-9, and inhibition of the expression of CX43 on astrocytes.Conclusion: Omarigliptin protects the integrity of the BBB in mice after ICH injury.Keywords: intracerebral hemorrhage, dipeptidyl peptidase, blood–brain barrier, omarigliptin

In this study, grand canonical Monte Carlo simulations (GCMC) and molecular dynamics simulations (MD) were used to construct models of activated carbon with hydroxyl-modified hexachlorobenzene basic unit contents of 0%, 12.5%, 25%, 35% and 50%. The mechanism of adsorption of carbon disulfide (CS2) by hydroxyl-modified activated carbon was then studied. It is found that the introduction of hydroxyl functional groups will improve the adsorption capacity of activated carbon for carbon disulfide. As far as the simulation results are concerned, the activated carbon model containing 25% hydroxyl modified activated carbon basic units has the best adsorption performance for carbon disulfide molecules at 318 K and atmospheric pressure. At the same time, the changes in the porosity, accessible surface area of the solvent, ultimate diameter and maximum pore diameter of the activated carbon model also led to great differences in the diffusion coefficient of carbon disulfide molecules in different hydroxyl-modified activated carbons. However, the same adsorption heat and temperature had little effect on the adsorption of carbon disulfide molecules.

Software specifications are essential for ensuring the reliability of software systems. Existing specification extraction approaches, however, suffer from limited generalizability and require manual efforts. We study the effectiveness of Large Language Models (LLMs) in generating software specifications from software documentation, utilizing Few-Shot Learning (FSL) to enable LLMs to generalize from a small number of examples. We compare the performance of LLMs with FSL to that of state-of-the-art specification extraction techniques and study the impact of prompt construction strategies on LLM performance. In addition, we conduct a comprehensive analysis of their symptoms and root causes of the failures to understand the pros and cons of LLMs and existing approaches. We also compare 11 LLMs' performance, cost, and response time for generating software specifications. Our findings include that (1) the best performing LLM outperforms existing approaches by 9.1--13.7% with a few similar examples, (2) the two dominant root causes combined (ineffective prompts and missing domain knowledge) result in 57--60% of LLM failures, and (3) most of the 11 LLMs achieve better or comparable performance compared to traditional techniques. Our study offers valuable insights for future research to improve specification generation.

Decompilation aims to recover the source code form of a binary executable. It has many applications in security and software engineering such as malware analysis, vulnerability detection and code reuse. A prominent challenge in decompilation is to recover variable names. We propose a novel method that leverages the synergy of large language model (LLM) and program analysis. Language models encode rich multi-modal knowledge, but its limited input size prevents providing sufficient global context for name recovery. We propose to divide the task to many LLM queries and use program analysis to correlate and propagate the query results, which in turn improves the performance of LLM by providing additional contextual information. Our results show that 75% of the recovered names are considered good by users and our technique outperforms the state-of-the-art technique by 16.5% and 20.23% in precision and recall, respectively.

To enhance the reliability and robustness of language identification (LID) and language diarization (LD) systems for heterogeneous populations and scenarios, there is a need for speech processing models to be trained on datasets that feature diverse language registers and speech patterns. We present the MERLIon CCS challenge, featuring a first-of-its-kind Zoom video call dataset of parent-child shared book reading, of over 30 hours with over 300 recordings, annotated by multilingual transcribers using a high-fidelity linguistic transcription protocol. The audio corpus features spontaneous and in-the-wild English-Mandarin code-switching, child-directed speech in non-standard accents with diverse language-mixing patterns recorded in a variety of home environments. This report describes the corpus, as well as LID and LD results for our baseline and several systems submitted to the MERLIon CCS challenge using the corpus., Accepted by Interspeech 2023, 5 pages, 2 figures, 3 tables

Reusing off-the-shelf code snippets from online repositories is a common practice, which significantly enhances the productivity of software developers. To find desired code snippets, developers resort to code search engines through natural language queries. Neural code search models are hence behind many such engines. These models are based on deep learning and gain substantial attention due to their impressive performance. However, the security aspect of these models is rarely studied. Particularly, an adversary can inject a backdoor in neural code search models, which return buggy or even vulnerable code with security/privacy issues. This may impact the downstream software (e.g., stock trading systems and autonomous driving) and cause financial loss and/or life-threatening incidents. In this paper, we demonstrate such attacks are feasible and can be quite stealthy. By simply modifying one variable/function name, the attacker can make buggy/vulnerable code rank in the top 11%. Our attack BADCODE features a special trigger generation and injection procedure, making the attack more effective and stealthy. The evaluation is conducted on two neural code search models and the results show our attack outperforms baselines by 60%. Our user study demonstrates that our attack is more stealthy than the baseline by two times based on the F1 score., Accepted to the 61st Annual Meeting of the Association for Computational Linguistics (ACL 2023)

We present view-synthesis autoencoders (VSA) in this paper, which is a self-supervised learning framework designed for vision transformers. Different from traditional 2D pretraining methods, VSA can be pre-trained with multi-view data. In each iteration, the input to VSA is one view (or multiple views) of a 3D object and the output is a synthesized image in another target pose. The decoder of VSA has several cross-attention blocks, which use the source view as value, source pose as key, and target pose as query. They achieve cross-attention to synthesize the target view. This simple approach realizes large-angle view synthesis and learns spatial invariant representation, where the latter is decent initialization for transformers on downstream tasks, such as 3D classification on ModelNet40, ShapeNet Core55, and ScanObjectNN. VSA outperforms existing methods significantly for linear probing and is competitive for fine-tuning. The code will be made publicly available., 13 pages, 12 figures

Self-supervised learning in computer vision trains on unlabeled data, such as images or (image, text) pairs, to obtain an image encoder that learns high-quality embeddings for input data. Emerging backdoor attacks towards encoders expose crucial vulnerabilities of self-supervised learning, since downstream classifiers (even further trained on clean data) may inherit backdoor behaviors from encoders. Existing backdoor detection methods mainly focus on supervised learning settings and cannot handle pre-trained encoders especially when input labels are not available. In this paper, we propose DECREE, the first backdoor detection approach for pre-trained encoders, requiring neither classifier headers nor input labels. We evaluate DECREE on over 400 encoders trojaned under 3 paradigms. We show the effectiveness of our method on image encoders pre-trained on ImageNet and OpenAI's CLIP 400 million image-text pairs. Our method consistently has a high detection accuracy even if we have only limited or no access to the pre-training dataset., Accepted at CVPR 2023. Code is available at https://github.com/GiantSeaweed/DECREE

In this paper, we propose a long-sequence modeling framework, named StreamPETR, for multi-view 3D object detection. Built upon the sparse query design in the PETR series, we systematically develop an object-centric temporal mechanism. The model is performed in an online manner and the long-term historical information is propagated through object queries frame by frame. Besides, we introduce a motion-aware layer normalization to model the movement of the objects. StreamPETR achieves significant performance improvements only with negligible computation cost, compared to the single-frame baseline. On the standard nuScenes benchmark, it is the first online multi-view method that achieves comparable performance (67.6% NDS & 65.3% AMOTA) with lidar-based methods. The lightweight version realizes 45.0% mAP and 31.7 FPS, outperforming the state-of-the-art method (SOLOFusion) by 2.3% mAP and 1.8x faster FPS. Code has been available at https://github.com/exiawsh/StreamPETR.git.

Long-term temporal fusion is a crucial but often overlooked technique in camera-based Bird's-Eye-View (BEV) 3D perception. Existing methods are mostly in a parallel manner. While parallel fusion can benefit from long-term information, it suffers from increasing computational and memory overheads as the fusion window size grows. Alternatively, BEVFormer adopts a recurrent fusion pipeline so that history information can be efficiently integrated, yet it fails to benefit from longer temporal frames. In this paper, we explore an embarrassingly simple long-term recurrent fusion strategy built upon the LSS-based methods and find it already able to enjoy the merits from both sides, i.e., rich long-term information and efficient fusion pipeline. A temporal embedding module is further proposed to improve the model's robustness against occasionally missed frames in practical scenarios. We name this simple but effective fusing pipeline VideoBEV. Experimental results on the nuScenes benchmark show that VideoBEV obtains leading performance on various camera-based 3D perception tasks, including object detection (55.4% mAP and 62.9% NDS), segmentation (48.6% vehicle mIoU), tracking (54.8% AMOTA), and motion prediction (0.80m minADE and 0.463 EPA). Code will be available.

Existing referring understanding tasks tend to involve the detection of a single text-referred object. In this paper, we propose a new and general referring understanding task, termed referring multi-object tracking (RMOT). Its core idea is to employ a language expression as a semantic cue to guide the prediction of multi-object tracking. To the best of our knowledge, it is the first work to achieve an arbitrary number of referent object predictions in videos. To push forward RMOT, we construct one benchmark with scalable expressions based on KITTI, named Refer-KITTI. Specifically, it provides 18 videos with 818 expressions, and each expression in a video is annotated with an average of 10.7 objects. Further, we develop a transformer-based architecture TransRMOT to tackle the new task in an online manner, which achieves impressive detection performance and outperforms other counterparts. The dataset and code will be available at https://github.com/wudongming97/RMOT., Accpeted by CVPR 2023. The dataset and code will be available at https://github.com/wudongming97/RMOT

Histone H4 lysine16 acetylation (H4K16Ac) modulates chromatin structure by serving as a switch from a repressive to a transcriptionally active state. This euchromatin mark is associated with active transcription. In this study, we investigated the effects of H4K16Ac on the expression of pro-fibrotic genes in lung fibroblasts from patients with idiopathic pulmonary fibrosis (IPF) and in an aging murine model of lung fibrosis. Methods: The lung tissues and fibroblasts from human IPF/non-IPF donors and from aged mice with/without bleomycin induced lung fibrosis were used in this study. The H4K16Ac levels were examined by immunohistochemistry or western blots. RNA silencing of H4K16Ac acetyltransferase Mof was used to reduce H4K16Ac levels in IPF fibroblasts. The effects of reduced H4K16Ac on pro-fibrotic gene expression were examined by western blots and real-time PCR. The association of H4K16Ac with these genes' promoter region were evaluated by ChIP assays. The gene expression profile in siRNA Mof transfected IPF cells were determined by RNA-Seq. The impact of H4K16Ac levels on lung fibrosis was evaluated in an aging murine model. Results: Aged mice with bleomycin induced lung fibrosis showed increased H4K16Ac levels. Human lung fibroblasts with siRNA Mof silencing demonstrated reduced H4K16Ac, and significantly down-regulated profibrotic genes, such as α-smooth muscle actin (α-SMA), collagen I, Nox4, and survivin. ChIP assays confirmed the associations of these pro-fibrotic genes' promoter region with H4K16Ac, while in siRNA Mof transfected cells the promoter/H4K16Ac associations were depleted. RNA-seq data demonstrated that Mof knockdown altered gene expression and cellular pathways, including cell damage and repair. In the aging mice model of persistent lung fibrosis, 18-month old mice given intra-nasal siRNA Mof from week 3 to 6 following bleomycin injury showed improved lung architecture, decreased total hydroxyproline content and lower levels of H4K16Ac. Conclusions: These results indicate a critical epigenetic regulatory role for histone H4K16Ac in the pathogenesis of pulmonary fibrosis, which will aid in the development of novel therapeutic strategies for age-related diseases such as IPF.

BackgroundIncreasing evidence supports that gut microbiota plays an important role in the development of cardiovascular diseases. The prevalence of sarcopenia is increasing in patients with heart failure. Muscle wasting is an independent predictor of death in heart failure patients.AimsIn this study, we aimed to explore the characteristics of gut microbiota and metabolites in heart failure patients with or without sarcopenia.MethodsFecal samples of 33 heart failure patients without sarcopenia, 29 heart failure patients with sarcopenia, and 15 controls were collected. The intestinal microbiota was analyzed using 16S rRNA sequencing and the metabolites were detected using the gas chromatography-mass spectrometry method.ResultsThere were significant differences in the overall microbial community structure and diversity between control and heart failure patients with or without sarcopenia. However, no clear clustering of samples was observed in heart failure with and without sarcopenia patients. Several bacterial, particularly Nocardiaceae, Pseudonocardiaceae, Alphaproteobacteria, and Slackia were significantly enriched in the heart failure patients without sarcopenia, while Synergistetes was more abundant in the heart failure patients with sarcopenia. Isobutyric acid, isovaleric acid, and valeric acid were lower in heart failure patients with sarcopenia than that without sarcopenia but lacked significance.ConclusionsThis study demonstrates that there are differences in the gut microbiota between control individuals and heart failure patients with or without sarcopenia. Modulating the gut microbiota may be a new target for the prevention and treatment of sarcopenia in heart failure patients.

Mainstream 3D representation learning approaches are built upon contrastive or generative modeling pretext tasks, where great improvements in performance on various downstream tasks have been achieved. However, we find these two paradigms have different characteristics: (i) contrastive models are data-hungry that suffer from a representation over-fitting issue; (ii) generative models have a data filling issue that shows inferior data scaling capacity compared to contrastive models. This motivates us to learn 3D representations by sharing the merits of both paradigms, which is non-trivial due to the pattern difference between the two paradigms. In this paper, we propose Contrast with Reconstruct (ReCon) that unifies these two paradigms. ReCon is trained to learn from both generative modeling teachers and single/cross-modal contrastive teachers through ensemble distillation, where the generative student guides the contrastive student. An encoder-decoder style ReCon-block is proposed that transfers knowledge through cross attention with stop-gradient, which avoids pretraining over-fitting and pattern difference issues. ReCon achieves a new state-of-the-art in 3D representation learning, e.g., 91.26% accuracy on ScanObjectNN. Codes have been released at https://github.com/qizekun/ReCon., Accepted at ICML 2023

Automated Program Repair (APR) improves software reliability by generating patches for a buggy program automatically. Recent APR techniques leverage deep learning (DL) to build models to learn to generate patches from existing patches and code corpora. While promising, DL-based APR techniques suffer from the abundant syntactically or semantically incorrect patches in the patch space. These patches often disobey the syntactic and semantic domain knowledge of source code and thus cannot be the correct patches to fix a bug. We propose a DL-based APR approach KNOD, which incorporates domain knowledge to guide patch generation in a direct and comprehensive way. KNOD has two major novelties, including (1) a novel three-stage tree decoder, which directly generates Abstract Syntax Trees of patched code according to the inherent tree structure, and (2) a novel domain-rule distillation, which leverages syntactic and semantic rules and teacher-student distributions to explicitly inject the domain knowledge into the decoding procedure during both the training and inference phases. We evaluate KNOD on three widely-used benchmarks. KNOD fixes 72 bugs on the Defects4J v1.2, 25 bugs on the QuixBugs, and 50 bugs on the additional Defects4J v2.0 benchmarks, outperforming all existing APR tools., This paper is accepted by 2023 IEEE/ACM 45th International Conference on Software Engineering (ICSE)

Monocular Depth Estimation (MDE) is a critical component in applications such as autonomous driving. There are various attacks against MDE networks. These attacks, especially the physical ones, pose a great threat to the security of such systems. Traditional adversarial training method requires ground-truth labels hence cannot be directly applied to self-supervised MDE that does not have ground-truth depth. Some self-supervised model hardening techniques (e.g., contrastive learning) ignore the domain knowledge of MDE and can hardly achieve optimal performance. In this work, we propose a novel adversarial training method for self-supervised MDE models based on view synthesis without using ground-truth depth. We improve adversarial robustness against physical-world attacks using L0-norm-bounded perturbation in training. We compare our method with supervised learning based and contrastive learning based methods that are tailored for MDE. Results on two representative MDE networks show that we achieve better robustness against various adversarial attacks with nearly no benign performance degradation., Initially accepted at ICLR2023 (Spotlight)

Deep Learning backdoor attacks have a threat model similar to traditional cyber attacks. Attack forensics, a critical counter-measure for traditional cyber attacks, is hence of importance for defending model backdoor attacks. In this paper, we propose a novel model backdoor forensics technique. Given a few attack samples such as inputs with backdoor triggers, which may represent different types of backdoors, our technique automatically decomposes them to clean inputs and the corresponding triggers. It then clusters the triggers based on their properties to allow automatic attack categorization and summarization. Backdoor scanners can then be automatically synthesized to find other instances of the same type of backdoor in other models. Our evaluation on 2,532 pre-trained models, 10 popular attacks, and comparison with 9 baselines show that our technique is highly effective. The decomposed clean inputs and triggers closely resemble the ground truth. The synthesized scanners substantially outperform the vanilla versions of existing scanners that can hardly generalize to different kinds of attacks.

Human-AI interactivity is a critical aspect that reflects the usability of multimodal large language models (MLLMs). However, existing end-to-end MLLMs only allow users to interact with them through language instructions, leading to the limitation of the interactive accuracy and efficiency. In this study, we present precise referring instructions that utilize diverse reference representations such as points and boxes as referring prompts to refer to the special region. This enables MLLMs to focus on the region of interest and achieve finer-grained interaction. Based on precise referring instruction, we propose ChatSpot, a unified end-to-end multimodal large language model that supports diverse forms of interactivity including mouse clicks, drag-and-drop, and drawing boxes, which provides a more flexible and seamless interactive experience. We also construct a multi-grained vision-language instruction-following dataset based on existing datasets and GPT-4 generating. Furthermore, we design a series of evaluation tasks to assess the effectiveness of region recognition and interaction. Experimental results showcase ChatSpot's promising performance., Comment: 15 pages, 8 figures

Multi-sensor fusion (MSF) is widely adopted for perception in autonomous vehicles (AVs), particularly for the task of 3D object detection with camera and LiDAR sensors. The rationale behind fusion is to capitalize on the strengths of each modality while mitigating their limitations. The exceptional and leading performance of fusion models has been demonstrated by advanced deep neural network (DNN)-based fusion techniques. Fusion models are also perceived as more robust to attacks compared to single-modal ones due to the redundant information in multiple modalities. In this work, we challenge this perspective with single-modal attacks that targets the camera modality, which is considered less significant in fusion but more affordable for attackers. We argue that the weakest link of fusion models depends on their most vulnerable modality, and propose an attack framework that targets advanced camera-LiDAR fusion models with adversarial patches. Our approach employs a two-stage optimization-based strategy that first comprehensively assesses vulnerable image areas under adversarial attacks, and then applies customized attack strategies to different fusion models, generating deployable patches. Evaluations with five state-of-the-art camera-LiDAR fusion models on a real-world dataset show that our attacks successfully compromise all models. Our approach can either reduce the mean average precision (mAP) of detection performance from 0.824 to 0.353 or degrade the detection score of the target object from 0.727 to 0.151 on average, demonstrating the effectiveness and practicality of our proposed attack framework.

Additional file 7: Table S5. Primers used in this study.

DETR has set up a simple end-to-end pipeline for object detection by formulating this task as a set prediction problem, showing promising potential. However, despite the significant progress in improving DETR, this paper identifies a problem of misalignment in the output distribution, which prevents the best-regressed samples from being assigned with high confidence, hindering the model's accuracy. We propose a metric, recall of best-regressed samples, to quantitively evaluate the misalignment problem. Observing its importance, we propose a novel Align-DETR that incorporates a localization precision-aware classification loss in optimization. The proposed loss, IA-BCE, guides the training of DETR to build a strong correlation between classification score and localization precision. We also adopt the mixed-matching strategy, to facilitate DETR-based detectors with faster training convergence while keeping an end-to-end scheme. Moreover, to overcome the dramatic decrease in sample quality induced by the sparsity of queries, we introduce a prime sample weighting mechanism to suppress the interference of unimportant samples. Extensive experiments are conducted with very competitive results reported. In particular, it delivers a 46 (+3.8)% AP on the DAB-DETR baseline with the ResNet-50 backbone and reaches a new SOTA performance of 50.2% AP in the 1x setting on the COCO validation set when employing the strong baseline DINO. Our code is available at https://github.com/FelixCaae/AlignDETR.

We develop, validate and apply a forward model to estimate stellar atmospheric parameters ($T_{\rm eff}$, $\log{g}$ and $\mathrm{[Fe/H]}$), revised distances and extinctions for 220 million stars with XP spectra from $\textit{Gaia}$ DR3. Instead of using $\textit{ab initio}$ stellar models, we develop a data-driven model of $\textit{Gaia}$ XP spectra as a function of the stellar parameters, with a few straightforward built-in physical assumptions. We train our model on stellar atmospheric parameters from the LAMOST survey, which provides broad coverage of different spectral types. We model the $\textit{Gaia}$ XP spectra with all of their covariances, augmented by 2MASS and WISE photometry that greatly reduces degeneracies between stellar parameters, yielding more precise determinations of temperature and dust reddening. Taken together, our approach overcomes a number of important limitations that the astrophysical parameters released in $\textit{Gaia}$ DR3 faced, and exploits the full information content of the data. We provide the resulting catalog of stellar atmospheric parameters, revised parallaxes and extinction estimates, with all their uncertainties. The modeling procedure also produces an estimate of the optical extinction curve at the spectral resolution of the XP spectra ($R \sim 20-100$), which agrees reasonably well with the ${R(V) = 3.1}$ CCM model. Remaining limitations that will be addressed in future work are that the model assumes a universal extinction law, ignores binary stars and does not cover all parts of the Hertzsprung-Russell Diagram ($\textit{e.g.}$, white dwarfs)., Comment: 27 pages, 22 figures, submitted to MNRAS, Data available at DOI: 10.5281/zenodo.7692680, v2.0: bugs in byline corrected

Additional file 1: Supplemental Table S1. Scoring standards for the disease activity index (DAI)1,2. Supplemental Table S2. Parameters and criteria of histological damage evaluation1,2. Supplemental Table S3. Primer pairs used in this experiment1. Supplemental Figure S1. Fecal bile acids (BAs) profiles of low birth weight (LBW) and normal birth weight (NBW) piglets 2 days after birth (n = 10). Supplemental Figure S2. Verification of depletion of the intestinal microbiota in mice (n = 8). Supplemental Figure S3. Microbial structure of NBW and LBW donor piglets and NBW-FMT and LBW-FMT recipient mice. Supplemental Figure S4. Microbial composition of NBW and LBW donor piglets and NBW-FMT and LBW-FMT recipient mice. Supplemental Figure S5. Colonic BA profiles of mice transplanted with the feces of LBW or NBW piglets (n = 8). Supplemental Figure S6. Effect of UDCA on inflammatory cytokines gene expression and macrophage polarization in J774A.1 cell.

Reverse engineering of protocol message formats is critical for many security applications. Mainstream techniques use dynamic analysis and inherit its low-coverage problem -- the inferred message formats only reflect the features of their inputs. To achieve high coverage, we choose to use static analysis to infer message formats from the implementation of protocol parsers. In this work, we focus on a class of extremely challenging protocols whose formats are described via constraint-enhanced regular expressions and parsed using finite-state machines. Such state machines are often implemented as complicated parsing loops, which are inherently difficult to analyze via conventional static analysis. Our new technique extracts a state machine by regarding each loop iteration as a state and the dependency between loop iterations as state transitions. To achieve high, i.e., path-sensitive, precision but avoid path explosion, the analysis is controlled to merge as many paths as possible based on carefully-designed rules. The evaluation results show that we can infer a state machine and, thus, the message formats, in five minutes with over 90% precision and recall, far better than state of the art. We also applied the state machines to enhance protocol fuzzers, which are improved by 20% to 230% in terms of coverage and detect ten more zero-days compared to baselines.

Although end-to-end multi-object trackers like MOTR enjoy the merits of simplicity, they suffer from the conflict between detection and association seriously, resulting in unsatisfactory convergence dynamics. While MOTRv2 partly addresses this problem, it demands an additional detection network for assistance. In this work, we serve as the first to reveal that this conflict arises from the unfair label assignment between detect queries and track queries during training, where these detect queries recognize targets and track queries associate them. Based on this observation, we propose MOTRv3, which balances the label assignment process using the developed release-fetch supervision strategy. In this strategy, labels are first released for detection and gradually fetched back for association. Besides, another two strategies named pseudo label distillation and track group denoising are designed to further improve the supervision for detection and association. Without the assistance of an extra detection network during inference, MOTRv3 achieves impressive performance across diverse benchmarks, e.g., MOT17, DanceTrack.

Due to the development of pre-trained language models, automated code generation techniques have shown great promise in recent years. However, the generated code is difficult to meet the syntactic constraints of the target language, especially in the case of Turducken-style code, where declarative code snippets are embedded within imperative programs. In this study, we summarize the lack of syntactic constraints into three significant challenges: (1) the efficient representation of syntactic constraints, (2) the effective integration of syntactic information, and (3) the scalable syntax-first decoding algorithm. To address these challenges, we propose a syntax-guided multi-task learning approach TurduckenGen. Specifically, we first explicitly append the type information to the code tokens to capture the representation of syntactic constraints. Then we formalize code generation with syntactic constraint representation as an auxiliary task to enable the model to learn the syntactic constraints of the code. Finally, the syntactically correct code is selected accurately from the multiple candidates with the help of the compiler feedback. Extensive experiments and comprehensive analysis demonstrate the effectiveness and general applicability of our approach after being compared with six state-of-the-art baselines on two Turducken-style code datasets. Finally, we conducted a human study and found the code quality generated by our approach is better than baselines in terms of code readability and semantic similarity., Comment: UNDER REVIEW

Additional file 1. Technical details such as mathematical derivations and the results of simulations with 10,000 SNPs.

A recent study has shown a phenomenon called neural collapse in that the within-class means of features and the classifier weight vectors converge to the vertices of a simplex equiangular tight frame at the terminal phase of training for classification. In this paper, we explore the corresponding structures of the last-layer feature centers and classifiers in semantic segmentation. Based on our empirical and theoretical analysis, we point out that semantic segmentation naturally brings contextual correlation and imbalanced distribution among classes, which breaks the equiangular and maximally separated structure of neural collapse for both feature centers and classifiers. However, such a symmetric structure is beneficial to discrimination for the minor classes. To preserve these advantages, we introduce a regularizer on feature centers to encourage the network to learn features closer to the appealing structure in imbalanced semantic segmentation. Experimental results show that our method can bring significant improvements on both 2D and 3D semantic segmentation benchmarks. Moreover, our method ranks 1st and sets a new record (+6.8% mIoU) on the ScanNet200 test leaderboard. Code will be available at https://github.com/dvlab-research/Imbalanced-Learning., Comment: Technical Report

In this paper, we propose a robust 3D detector, named Cross Modal Transformer (CMT), for end-to-end 3D multi-modal detection. Without explicit view transformation, CMT takes the image and point clouds tokens as inputs and directly outputs accurate 3D bounding boxes. The spatial alignment of multi-modal tokens is performed by encoding the 3D points into multi-modal features. The core design of CMT is quite simple while its performance is impressive. It achieves 74.1\% NDS (state-of-the-art with single model) on nuScenes test set while maintaining faster inference speed. Moreover, CMT has a strong robustness even if the LiDAR is missing. Code is released at https://github.com/junjie18/CMT.

Additional file 5: Table S4. Predicted interacting proteins of G. hirsutum PRC2 core components in the ccNET database.

Additional file 1: Table S1. Identities between Arabidopsis and cotton PRC2 core components.

Additional file 5: Table S4. Predicted interacting proteins of G. hirsutum PRC2 core components in the ccNET database.

Additional file 2: Table S2. Predicted domain organization of G. hirsutum PRC2 proteins.

3D object detectors usually rely on hand-crafted proxies, e.g., anchors or centers, and translate well-studied 2D frameworks to 3D. Thus, sparse voxel features need to be densified and processed by dense prediction heads, which inevitably costs extra computation. In this paper, we instead propose VoxelNext for fully sparse 3D object detection. Our core insight is to predict objects directly based on sparse voxel features, without relying on hand-crafted proxies. Our strong sparse convolutional network VoxelNeXt detects and tracks 3D objects through voxel features entirely. It is an elegant and efficient framework, with no need for sparse-to-dense conversion or NMS post-processing. Our method achieves a better speed-accuracy trade-off than other mainframe detectors on the nuScenes dataset. For the first time, we show that a fully sparse voxel-based representation works decently for LIDAR 3D object detection and tracking. Extensive experiments on nuScenes, Waymo, and Argoverse2 benchmarks validate the effectiveness of our approach. Without bells and whistles, our model outperforms all existing LIDAR methods on the nuScenes tracking test benchmark., Comment: In CVPR 2023, Code and models are available at https://github.com/dvlab-research/VoxelNeXt

Additional file 7: Table S5. Primers used in this study.

Inferring protocol formats is critical for many security applications. However, existing format-inference techniques often miss many formats, because almost all of them are in a fashion of dynamic analysis and rely on a limited number of network packets to drive their analysis. If a feature is not present in the input packets, the feature will be missed in the resulting formats. We develop a novel static program analysis for format inference. It is well-known that static analysis does not rely on any input packets and can achieve high coverage by scanning every piece of code. However, for efficiency and precision, we have to address two challenges, namely path explosion and disordered path constraints. To this end, our approach uses abstract interpretation to produce a novel data structure called the abstract format graph. It delimits precise but costly operations to only small regions, thus ensuring precision and efficiency at the same time. Our inferred formats are of high coverage and precisely specify both field boundaries and semantic constraints among packet fields. Our evaluation shows that we can infer formats for a protocol in one minute with >95% precision and recall, much better than four baseline techniques. Our inferred formats can substantially enhance existing protocol fuzzers, improving the coverage by 20% to 260% and discovering 53 zero-days with 47 assigned CVEs. We also provide case studies of adopting our inferred formats in other security applications including traffic auditing and intrusion detection.

Efficiency is quite important for 3D lane detection due to practical deployment demand. In this work, we propose a simple, fast, and end-to-end detector that still maintains high detection precision. Specifically, we devise a set of fully convolutional heads based on row-wise classification. In contrast to previous counterparts, ours supports recognizing both vertical and horizontal lanes. Besides, our method is the first one to perform row-wise classification in bird-eye-view. In the heads, we split feature into multiple groups and every group of feature corresponds to a lane instance. During training, the predictions are associated with lane labels using the proposed single-win one-to-one matching to compute loss, and no post-processing operation is demanded for inference. In this way, our proposed fully convolutional detector, GroupLane, realizes end-to-end detection like DETR. Evaluated on 3 real world 3D lane benchmarks, OpenLane, Once-3DLanes, and OpenLane-Huawei, GroupLane adopting ConvNext-Base as the backbone outperforms the published state-of-the-art PersFormer by 13.6% F1 score in the OpenLane validation set. Besides, GroupLane with ResNet18 still surpasses PersFormer by 4.9% F1 score, while the inference speed is nearly 7x faster and the FLOPs is only 13.3% of it.

Supplemental 1

Smart homes contain diverse sensors and actuators controlled by IoT apps that provide custom automation. Prior works showed that an adversary could exploit physical interaction vulnerabilities among apps and put the users and environment at risk, e.g., to break into a house, an adversary turns on the heater to trigger an app that opens windows when the temperature exceeds a threshold. Currently, the safe behavior of physical interactions relies on either app code analysis or dynamic analysis of device states with manually derived policies by developers. However, existing works fail to achieve sufficient breadth and fidelity to translate the app code into their physical behavior or provide incomplete security policies, causing poor accuracy and false alarms. In this paper, we introduce a new approach, IoTSeer, which efficiently combines app code analysis and dynamic analysis with new security policies to discover physical interaction vulnerabilities. IoTSeer works by first translating sensor events and actuator commands of each app into a physical execution model (PeM) and unifying PeMs to express composite physical execution of apps (CPeM). CPeM allows us to deploy IoTSeer in different smart homes by defining its execution parameters with minimal data collection. IoTSeer supports new security policies with intended/unintended physical channel labels. It then efficiently checks them on the CPeM via falsification, which addresses the undecidability of verification due to the continuous and discrete behavior of IoT devices. We evaluate IoTSeer in an actual house with 14 actuators, six sensors, and 39 apps. IoTSeer discovers 16 unique policy violations, whereas prior works identify only 2 out of 16 with 18 falsely flagged violations. IoTSeer only requires 30 mins of data collection for each actuator to set the CPeM parameters and is adaptive to newly added, removed, and relocated devices., Comment: Published in ACM CCS 2022

We present the analysis of seven microlensing planetary events with planet/host mass ratios $q < 10^{-4}$: KMT-2017-BLG-1194, KMT-2017-BLG-0428, KMT-2019-BLG-1806, KMT-2017-BLG-1003, KMT-2019-BLG-1367, OGLE-2017-BLG-1806, and KMT-2016-BLG-1105. They were identified by applying the Korea Microlensing Telescope Network (KMTNet) AnomalyFinder algorithm to 2016--2019 KMTNet events. A Bayesian analysis indicates that all the lens systems consist of a cold super-Earth orbiting an M or K dwarf. Together with 17 previously published and three that will be published elsewhere, AnomalyFinder has found a total of 27 planets that have solutions with $q < 10^{-4}$ from 2016--2019 KMTNet events, which lays the foundation for the first statistical analysis of the planetary mass-ratio function based on KMTNet data. By reviewing the 27 planets, we find that the missing planetary caustics problem in the KMTNet planetary sample has been solved by AnomalyFinder. We also find a desert of high-magnification planetary signals ($A \gtrsim 65$), and a follow-up project for KMTNet high-magnification events could detect at least two more $q < 10^{-4}$ planets per year and form an independent statistical sample., Comment: AJ in press

Grid-interactive building control is a challenging and important problem for reducing carbon emissions, increasing energy efficiency, and supporting the electric power grid. Currently researchers and practitioners are confronted with a choice of control strategies ranging from model-free (purely data-driven) to model-based (directly incorporating physical knowledge) to hybrid methods that combine data and models. In this work, we identify state-of-the-art methods that span this methodological spectrum and evaluate their performance for multi-zone building HVAC control in the context of three demand response programs. We demonstrate, in this context, that hybrid methods offer many benefits over both purely model-free and model-based methods as long as certain requirements are met. In particular, hybrid controllers are relatively sample efficient, fast online, and high accuracy so long as the test case falls within the distribution of training data. Like all data-driven methods, hybrid controllers are still subject to generalization errors when applied to out-of-sample scenarios. Key takeaways for control strategies are summarized and the developed software framework is open-sourced.

Lyrics recognition is an important task in music processing. Despite traditional algorithms such as the hybrid HMM- TDNN model achieving good performance, studies on applying end-to-end models and self-supervised learning (SSL) are limited. In this paper, we first establish an end-to-end baseline for lyrics recognition and then explore the performance of SSL models on lyrics recognition task. We evaluate a variety of upstream SSL models with different training methods (masked reconstruction, masked prediction, autoregressive reconstruction, and contrastive learning). Our end-to-end self-supervised models, evaluated on the DAMP music dataset, outperform the previous state-of-the-art (SOTA) system by 5.23% for the dev set and 2.4% for the test set even without a language model trained by a large corpus. Moreover, we investigate the effect of background music on the performance of self-supervised learning models and conclude that the SSL models cannot extract features efficiently in the presence of background music. Finally, we study the out-of-domain generalization ability of the SSL features considering that those models were not trained on music datasets., 4 pages, 2 figures, 3 tables

Transient recovery characteristics of DC voltage are critical to analyze the stability criterion and the fault ride-through capability of voltage source converter (VSC)-based DC power systems. In this paper, the dynamic model of a flexible DC power grid was initially established, and then, the dynamic operation mechanism of the system was presented by analyzing the movement locus of the operating point when disturbances occur in the DC grid. Subsequently, the DC transient voltage limit and the DC fault clearing time were calculated in order to obtain the transient stability criterion of the DC power system. In addition, a novel voltage transient recovery control strategy was proposed by improving the droop control of the DC voltage to extend the transient stability margin of the power system. Finally, a typical multi-terminal flexible DC power grid was developed on a hardware-in-the-loop platform. The results demonstrated that the voltage recovery ability of the DC power system can be quantitatively estimated by the stability criterion, while the system transient stability can be improved by the proposed control strategy.

Input constraints are useful for many software development tasks. For example, input constraints of a function enable the generation of valid inputs, i.e., inputs that follow these constraints, to test the function deeper. API functions of deep learning (DL) libraries have DL specific input constraints, which are described informally in the free form API documentation. Existing constraint extraction techniques are ineffective for extracting DL specific input constraints. To fill this gap, we design and implement a new technique, DocTer, to analyze API documentation to extract DL specific input constraints for DL API functions. DocTer features a novel algorithm that automatically constructs rules to extract API parameter constraints from syntactic patterns in the form of dependency parse trees of API descriptions. These rules are then applied to a large volume of API documents in popular DL libraries to extract their input parameter constraints. To demonstrate the effectiveness of the extracted constraints, DocTer uses the constraints to enable the automatic generation of valid and invalid inputs to test DL API functions. Our evaluation on three popular DL libraries (TensorFlow, PyTorch, and MXNet) shows that the precision of DocTer in extracting input constraints is 85.4%. DocTer detects 94 bugs from 174 API functions, including one previously unknown security vulnerability that is now documented in the CVE database, while a baseline technique without input constraints detects only 59 bugs. Most (63) of the 94 bugs are previously unknown, 54 of which have been fixed or confirmed by developers after we report them. In addition, DocTer detects 43 inconsistencies in documents, 39 of which are fixed or confirmed., This paper is accepted by the 31st ACM SIGSOFT International Symposium on Software Testing and Analysis (ISSTA) in 2022