Traditional object detection systems are typically constrained to predefined categories, limiting their applicability in dynamic environments. In contrast, open-vocabulary object detection (OVD) enables the identification of objects from novel classes not present in the training set. Recent advances in visual-language modeling have led to significant progress of OVD. However, prior works face challenges in either adapting the single-scale image backbone from CLIP to the detection framework or ensuring robust visual-language alignment. We propose Visual-Language Detection (VLDet), a novel framework that revamps feature pyramid for fine-grained visual-language alignment, leading to improved OVD performance. With the VL-PUB module, VLDet effectively exploits the visual-language knowledge from CLIP and adapts the backbone for object detection through feature pyramid. In addition, we introduce the SigRPN block, which incorporates a sigmoid-based anchor-text contrastive alignment loss to improve detection of novel categories. Through extensive experiments, our approach achieves 58.7 AP for novel classes on COCO2017 and 24.8 AP on LVIS, surpassing all state-of-the-art methods and achievin
Weakly interacting massive particles (WIMPs) constitute a paradigm in the search for particle dark matter. In contrast to supersymmetry (SUSY), we explore the possibility that WIMP dark-matter acts as mediator of neutrino mass generation. We examine in detail the phenomenology of fermionic dark matter in the revamped (or singlet-triplet) scotogenic model and study its collider implications. Unlike SUSY WIMP dark-matter, collider searches for the Lightest Scotogenic Particle (LSP) at LHC/LHC-HL are strongly complementary to charged lepton flavor violation probes and dark matter studies.
This paper introduces a novel conformal selection procedure, inspired by the Neyman--Pearson paradigm, to maximize the power of selecting qualified units while maintaining false discovery rate (FDR) control. Existing conformal selection methods may yield suboptimal power due to their reliance on conformal p-values, which are derived by substituting unobserved future outcomes with thresholds set by the null hypothesis. This substitution invalidates the exchangeability between imputed nonconformity scores for test data and those derived from calibration data, resulting in reduced power. In contrast, our approach circumvents the need for conformal p-values by constructing a likelihood-ratio-based decision rule that directly utilizes observed covariates from both calibration and test samples. The asymptotic optimality and FDR control of the proposed method are established under a correctly specified model, and modified selection procedures are introduced to improve power under model misspecification. The proposed methods are computationally efficient and can be readily extended to handle covariate shifts, making them well-suited for real-world applications. Simulation results show that
Implementing the axion concept in the context of 3-3-1 extensions of the Standard Model (SM) leads to richer properties than in the simpler axion setups, and related to the Dirac neutrino seesaw mechanism. In this way the smallness of neutrino masses, the strong CP problem, the nature of dark matter and the number of families all have a common origin. Besides having an enhanced coupling to photons, our revamped axion can also be distinguished from DFSZ and KSVZ axions through its couplings to fermions. The latter lead to interesting phenomenological consequences, including flavor-changing axion-emitting two-body K, B and D meson decays.
This exploratory field study investigates the integration of innovative forms of recitation tasks in a first-year introductory mechanics course, focusing on smartphone-based experimental tasks alongside programming and standard recitation tasks. Smartphones, combined with external sensor modules, serve as a gateway enabling students to conduct various low-cost and authentic physics experiments with first-hand data collection outside traditional lab settings. These tasks aim to enhance students' agency in independent physics experimentation and enrich homework assignments by dissolving boundaries between lectures, recitation sessions, and traditional labs, and thereby linking theoretical and experimental aspects of undergraduate physics education. To explore this potential, we implemented and evaluated a sample set of nine smartphone-based experimental tasks and, for comparison, three programming tasks as weekly exercises in a first-year physics course at RWTH Aachen University. We investigated students' perceptions of learning with these new tasks through twelve short surveys involving up to 188 participants. In two additional surveys with 108 and 78 participants, students assessed
Recent approaches attempt to adapt powerful interactive segmentation models, such as SAM, to interactive matting and fine-tune the models based on synthetic matting datasets. However, models trained on synthetic data fail to generalize to complex and occlusion scenes. We address this challenge by proposing a new matting dataset based on the COCO dataset, namely COCO-Matting. Specifically, the construction of our COCO-Matting includes accessory fusion and mask-to-matte, which selects real-world complex images from COCO and converts semantic segmentation masks to matting labels. The built COCO-Matting comprises an extensive collection of 38,251 human instance-level alpha mattes in complex natural scenarios. Furthermore, existing SAM-based matting methods extract intermediate features and masks from a frozen SAM and only train a lightweight matting decoder by end-to-end matting losses, which do not fully exploit the potential of the pre-trained SAM. Thus, we propose SEMat which revamps the network architecture and training objectives. For network architecture, the proposed feature-aligned transformer learns to extract fine-grained edge and transparency features. The proposed matte-ali
The KNT(W) data-driven determinations of the hadronic vacuum polarization (HVP) are crucial inputs to previous and future Standard Model (SM) predictions of the muon's anomalous magnetic moment, $a_μ$. With the muon $g$$-$$2$'s new physics case uncertain due to disagreeing HVP evaluations, new SM predictions and experimental measurements of $a_μ$ expected soon, and a complete revamp of the KNTW analysis framework underway, this letter motivates and describes a blinding scheme for data-driven HVP determinations that has been implemented for future KNTW analyses.
Interesting data on Gamma Ray Burts (GRBs) and Cosmic Rays (CRs) have recently been made public. GRB221009A has a record ``peak energy". The CR electron spectrum has been measured to unprecedented high energies and exhibits a ``knee" akin to the ones in all-particle or individual-element CR nuclei. IceCube has not seen high-energy neutrinos associated with GRBs. AMS has published a CR positron spectrum conducive to much speculation. We examine these data in the light of the ``CannonBall Model" of GRBs and CRs, in which they are intimately related and which they do strongly validate.
In the arena of privacy-preserving machine learning, differentially private stochastic gradient descent (DP-SGD) has outstripped the objective perturbation mechanism in popularity and interest. Though unrivaled in versatility, DP-SGD requires a non-trivial privacy overhead (for privately tuning the model's hyperparameters) and a computational complexity which might be extravagant for simple models such as linear and logistic regression. This paper revamps the objective perturbation mechanism with tighter privacy analyses and new computational tools that boost it to perform competitively with DP-SGD on unconstrained convex generalized linear problems.
Deep Learning models, such as those used in an autonomous vehicle are vulnerable to adversarial attacks where an attacker could place an adversarial object in the environment, leading to mis-classification. Generating these adversarial objects in the digital space has been extensively studied, however successfully transferring these attacks from the digital realm to the physical realm has proven challenging when controlling for real-world environmental factors. In response to these limitations, we introduce REVAMP, an easy-to-use Python library that is the first-of-its-kind tool for creating attack scenarios with arbitrary objects and simulating realistic environmental factors, lighting, reflection, and refraction. REVAMP enables researchers and practitioners to swiftly explore various scenarios within the digital realm by offering a wide range of configurable options for designing experiments and using differentiable rendering to reproduce physically plausible adversarial objects. We will demonstrate and invite the audience to try REVAMP to produce an adversarial texture on a chosen object while having control over various scene parameters. The audience will choose a scene, an obj
The surge in developing deep learning models for diagnosing skin lesions through image analysis is notable, yet their clinical black faces challenges. Current dermatology AI models have limitations: limited number of possible diagnostic outputs, lack of real-world testing on uncommon skin lesions, inability to detect out-of-distribution images, and over-reliance on dermoscopic images. To address these, we present an All-In-One \textbf{H}ierarchical-\textbf{O}ut of Distribution-\textbf{C}linical Triage (HOT) model. For a clinical image, our model generates three outputs: a hierarchical prediction, an alert for out-of-distribution images, and a recommendation for dermoscopy if clinical image alone is insufficient for diagnosis. When the recommendation is pursued, it integrates both clinical and dermoscopic images to deliver final diagnosis. Extensive experiments on a representative cutaneous lesion dataset demonstrate the effectiveness and synergy of each component within our framework. Our versatile model provides valuable decision support for lesion diagnosis and sets a promising precedent for medical AI applications.
Recently advanced non-geostationary (NGSO) satellite networks represented by large constellations and advanced payloads provide great promises for enabling high-quality Internet connectivity to any place on Earth. However, the traditional approach to satellite operations cannot address the new challenges in the NGSO satellite networks imposed by the significant increase in complexity, security, resilience, and environmental concerns. Therefore, a reliable, sustainable, and efficient approach is required for the entire life-cycle of satellite network operations. This paper provides a timely response to the new challenges and proposes a novel approach called "SatAIOps" as an overall solution. Through our discussion on the current challenges of the advanced satellite networks, SatAIOps and its functional modules in the entire life-cycle of satellites are proposed, with some example technologies given. SatAIOps provides a new perspective for addressing operational challenges with trustworthy and responsible AI technologies. It enables a new framework for evolving and collaborative efforts from research and industry communities.
In the present work, our main objective is to investigate the orbits of spinning test particles around a Schwarzschild black hole under the influence of a quintessence matter field (SQBH). We begin with the dynamics of the spinning test particles around SQBH which is governed by the Mathisson-Papapetrou-Dixon (MPD) equations under the pole-dipole approximation, where the gravitational field and the higher multipoles of the particle are neglected. Depending on the types of saddle points,the effective potential are classified and the possibility of chaotic orbits is discussed. The inner most stable circular orbits (ISCOs) of the spinning particle around SQBH are addressed, as are the effects of the parameters $S$ (particles' spin) and $ε$ (equation of state parameter). Later, Periastron precession is investigated up to the first-order spin correction for a spinning particle moving in nearly circular orbits around SQBH. It is noted that the addition of particle's spin revamps the results obtained for the non-spinning particles and also articulates the some interesting observational properties of the SQBH. Additionally, we discuss the ramifications of employing first-order spin correct
Adversarial Propagation (AdvProp) is an effective way to improve recognition models, leveraging adversarial examples. Nonetheless, AdvProp suffers from the extremely slow training speed, mainly because: a) extra forward and backward passes are required for generating adversarial examples; b) both original samples and their adversarial counterparts are used for training (i.e., 2$\times$ data). In this paper, we introduce Fast AdvProp, which aggressively revamps AdvProp's costly training components, rendering the method nearly as cheap as the vanilla training. Specifically, our modifications in Fast AdvProp are guided by the hypothesis that disentangled learning with adversarial examples is the key for performance improvements, while other training recipes (e.g., paired clean and adversarial training samples, multi-step adversarial attackers) could be largely simplified. Our empirical results show that, compared to the vanilla training baseline, Fast AdvProp is able to further model performance on a spectrum of visual benchmarks, without incurring extra training cost. Additionally, our ablations find Fast AdvProp scales better if larger models are used, is compatible with existing da
This research aims to improve the user interface of SIAK-NG (Next Generation Academic Information System), the academic portal website of the University of Indonesia through a design thinking approach. Despite being in existence for several decades and undergoing multiple improvements, there has been no specific effort to evaluate the quality of the user interface. In line with the revamping of SIAK-NG according to the University of Indonesia 2019-2024 master plan, this study aims to provide the necessary insights. The research focuses on redesigning the website interface of SIAK-NG to address user complaints and difficulties, particularly related to the unsatisfactory interface design. The Design Thinking approach is employed to generate solutions that meet the needs of active University of Indonesia students who are the primary users. Through methods such as storyboarding, empathy mapping, usability testing, and others, the author will design recommendations that align with user requirements.
ESPnet-ST-v2 is a revamp of the open-source ESPnet-ST toolkit necessitated by the broadening interests of the spoken language translation community. ESPnet-ST-v2 supports 1) offline speech-to-text translation (ST), 2) simultaneous speech-to-text translation (SST), and 3) offline speech-to-speech translation (S2ST) -- each task is supported with a wide variety of approaches, differentiating ESPnet-ST-v2 from other open source spoken language translation toolkits. This toolkit offers state-of-the-art architectures such as transducers, hybrid CTC/attention, multi-decoders with searchable intermediates, time-synchronous blockwise CTC/attention, Translatotron models, and direct discrete unit models. In this paper, we describe the overall design, example models for each task, and performance benchmarking behind ESPnet-ST-v2, which is publicly available at https://github.com/espnet/espnet.
Keyword extraction is a task of text mining. It is applied to increase search volume in SEO and ads. Implemented in auto-tagging, it makes tagging on a mass scale of online articles and photos efficiently and accurately. BAT is invented for auto-tagging which served as awoo's AI marketing platform (AMP). awoo AMP not only provides service as a customized recommender system but also increases the converting rate in E-commerce. The strength of BAT converges faster and better than other SOTA models, as its 4-layer structure achieves the best F scores at 50 epochs. In other words, it performs better than other models which require deeper layers at 100 epochs. To generate rich and clean tags, awoo creates new objective functions to maintain similar ${\rm F_1}$ scores with cross-entropy while enhancing ${\rm F_2}$ scores simultaneously. To assure the even better performance of F scores awoo revamps the learning rate strategy proposed by Transformer \cite{Transformer} to increase ${\rm F_1}$ and ${\rm F_2}$ scores at the same time.
We optimize and characterize the preparation of 3-trimethoxysilyl propylmethacrylate (TPM) colloidal suspensions for three-dimensional confocal microscopy. We revisit a simple synthesis of TPM microspheres by nucleation of droplets from pre-hydrolyzed TPM oil in a 'zero-flow' regime, and demonstrate how precise and reproducible control of particle size may be achieved via single-step nucleation with a focus on how the reagents are mixed. We also revamp the conventional dyeing method for TPM particles to achieve uniform transfer of a fluorophore to the organosilica droplets, improving particle identification. Finally, we illustrate how a ternary mixture of tetralin, trichloroethylene and tetrachloroethylene may be used as a suspension medium which matches the refractive index of these particles while allowing independent control of the density mismatch between particle and solvent.
Despite the fact that multilingual agreement (MA) has shown its importance for multilingual neural machine translation (MNMT), current methodologies in the field have two shortages: (i) require parallel data between multiple language pairs, which is not always realistic and (ii) optimize the agreement in an ambiguous direction, which hampers the translation performance. We present \textbf{B}idirectional \textbf{M}ultilingual \textbf{A}greement via \textbf{S}witched \textbf{B}ack-\textbf{t}ranslation (\textbf{BMA-SBT}), a novel and universal multilingual agreement framework for fine-tuning pre-trained MNMT models, which (i) exempts the need for aforementioned parallel data by using a novel method called switched BT that creates synthetic text written in another source language using the translation target and (ii) optimizes the agreement bidirectionally with the Kullback-Leibler Divergence loss. Experiments indicate that BMA-SBT clearly improves the strong baselines on the task of MNMT with three benchmarks: TED Talks, News, and Europarl. In-depth analyzes indicate that BMA-SBT brings additive improvements to the conventional BT method.