Automatic image captioning is a promising technique for conveying visual information using natural language. It can benefit various tasks in satellite remote sensing, such as environmental monitoring, resource management, disaster management, etc. However, one of the main challenges in this domain is the lack of large-scale image-caption datasets, as they require a lot of human expertise and effort to create. Recent research on large language models (LLMs) has demonstrated their impressive performance in natural language understanding and generation tasks. Nonetheless, most of them cannot handle images (GPT-3.5, Falcon, Claude, etc.), while conventional captioning models pre-trained on general ground-view images often fail to produce detailed and accurate captions for aerial images (BLIP, GIT, CM3, CM3Leon, etc.). To address this problem, we propose a novel approach: Automatic Remote Sensing Image Captioning (ARSIC) to automatically collect captions for remote sensing images by guiding LLMs to describe their object annotations. We also present a benchmark model that adapts the pre-trained generative image2text model (GIT) to generate high-quality captions for remote-sensing images. Our evaluation demonstrates the effectiveness of our approach for collecting captions for remote sensing images.
This paper provides a comprehensive and high level technical overview of BC technology by creating a common language through business and technology about BC. Then an overview of a BC technology, its architectures, classification, challenges, etc. is presented. Finally, some recommendations are provided to re-searchers that will have to be tackled before deploying the next generation of IoT applications based on BC technology.
It has been recently shown in the literature that the sample averages from online learning experiments are biased when used to estimate the mean reward. To correct the bias, off-policy evaluation methods, including importance sampling and doubly robust estimators, typically calculate the conditional propensity score, which is ill-defined for non-randomized policies such as UCB. This paper provides a procedure to debias the samples using bootstrap, which doesn't require the knowledge of the reward distribution and can be applied to any adaptive policies. Numerical experiments demonstrate the effective bias reduction for samples generated by popular multi-armed bandit algorithms such as Explore-Then-Commit (ETC), UCB, Thompson sampling (TS) and $ε$-greedy (EG). We analyze and provide theoretical justifications for the procedure under the ETC algorithm, including the asymptotic convergence of the bias decay rate in the real and bootstrap worlds.
Francesco Liberati, Andrea Tortorelli, Cesar Mazquiaran
et al.
This paper discusses the problem of assembly line control and introduces an optimal control formulation that can be used to improve the performance of the assembly line, in terms of cycle time minimization, resources' utilization, etc. A deterministic formulation of the problem is introduced, based on mixed-integer linear programming. A simple numerical simulation provides a first proof of the proposed concept.
Recent evidence shows that convolutional neural networks (CNNs) are biased towards textures so that CNNs are non-robust to adversarial perturbations over textures, while traditional robust visual features like SIFT (scale-invariant feature transforms) are designed to be robust across a substantial range of affine distortion, addition of noise, etc with the mimic of human perception nature. This paper aims to leverage good properties of SIFT to renovate CNN architectures towards better accuracy and robustness. We borrow the scale-space extreme value idea from SIFT, and propose extreme value preserving networks (EVPNets). Experiments demonstrate that EVPNets can achieve similar or better accuracy than conventional CNNs, while achieving much better robustness on a set of adversarial attacks (FGSM,PGD,etc) even without adversarial training.
We study several variants of Euler sums by using the methods of contour integration and residue theorem. These variants exhibit nice properties such as closed forms, reduction, etc., like classical Euler sums. In addition, we also define a variant of multiple zeta values of level 2, and give some identities on relations between these variants of Euler sums and the variant of multiple zeta values.
Aditi Vinay Chandak, Surekha Dubey Godbole, Tanvi Balwani
et al.
Speech is considered as a basic fundamental means of communication, which makes the human being superior than other forms of life. Speech and language therapist judgement of speech is consider as the most perfect because the assessment is mainly subjective and it depends on the perception of individual. This will involve both assessment of the intelligibility and quality of the patient’s speech, and observation of the visible aspects of articulation. The best way is to use perpetual assessment, to highlight potential areas of difficulty, then objective, instrumental assessment of these areas, to confirm the nature and severity of their involvement. Correlation of the vocal signal changes with the characteristics of the prosthesis and the specific types of errors in the prosthetic act would be an essential achievement in the way of improving the outcome of the prosthetic action. It is the responsibility of the prosthodontist to construct dentures as accurately as possible, so as to improve speech sound production with dentures, minimize the period of adaptation and thereby, increase the self-confidence of the patient. For this its very important to have knowledge about assessment of speech. Since past many years clinicians have faced problems in assessing speech. In this article clinical application of speech test in relation to complete dentures have been highlighted.
As we age, our hearts undergo changes which result in reduction in complexity of physiological interactions between different control mechanisms. This results in a potential risk of cardiovascular diseases which are the number one cause of death globally. Since cardiac signals are nonstationary and nonlinear in nature, complexity measures are better suited to handle such data. In this study, non-invasive methods for detection of cardiac aging using complexity measures are explored. Lempel-Ziv (LZ) complexity, Approximate Entropy (ApEn) and Effort-to-Compress (ETC) measures are used to differentiate between healthy young and old subjects using heartbeat interval data. We show that both LZ and ETC complexity measures are able to differentiate between young and old subjects with only 10 data samples while ApEn requires at least 15 data samples.
In the $N$-body problem, a simple choreography is a periodic solution, where all masses chase each other on a single loop. In this paper we prove that for the planar Newtonian $N$-body problem with equal masses, $N \ge 3$, there are at least $2^{N-3} + 2^{[(N-3)/2]}$ different main simple choreographies. This confirms a conjecture given by Chenciner and etc. in \cite{CGMS02}.
The paper establishes error orders for integral limit approximations to the traces of products of Toeplitz matrices generated by integrable real symmetric functions defined on the unit circle. These approximations and the corresponding error bounds are of importance in the statistical analysis of discrete-time stationary processes: asymptotic distributions and large deviations of Toeplitz type random quadratic forms, estimation of the spectral parameters and functionals, etc.
The paper covers the new model of wage distribution in typical group of people. The model provides the opportunity to reparameterize applicable income distribution model: Pareto, logarithmically normal, logarithmically logistic, Dagum etc. The model ensures the graduation of Gini index values by polynomial degree of wage distribution as well as different types of income distribution. The given approach clarifies the nature of income inequality.
In this paper, a new method for generation of infinite series of symmetric identities written for exponential sums in real numbers is proposed. Such systems have numerous applications in theory of numbers, chaos theory, algorithmic complexity, dynamic systems, etc. Properties of generated identities are studied. Relations of the introduced method for generation of symmetric exponential sums to the Morse-Hedlund sequence and to the theory of magic squares are established.
General Relativity offers the possibility to model attributes of matter, like mass, momentum, angular momentum, spin, chirality etc. from pure space, endowed only with a single field that represents its Riemannian geometry. I review this picture of `Geometrodynamics' and comment on various developments after Einstein.
The Faulkes Telescope (FT) Project is an educational and research arm of the Las Cumbres Observatory Global Telescope Network (LCOGTN). As well as producing spectacular images of galaxies, nebulae, supernovae remnants, star clusters, etc., the FT team is involved in several projects pursuing scientific goals. Many of these projects also incorporate data collected and analysed by schools and amateur astronomers.
Matter collapsing to a singularity in a gravitational field is still an intriguing question. Similar situation arises when discussing the very early universe or a universe recollapsing to a singularity. It is suggested that inclusion of mutual gravitational interactions among the collapsing particles can avert a singularity and give finite value for various physical quantities like entropy, density, etc.
We review various inequalities on the order and the spacing of energy levels, wave function at the origin, etc... which were obtained since 1977 in the framework of the Schrodinger equation and applied to quarkonium and also to muonic atoms and alcaline atoms. We also present a fit of mesons and baryons made of b, c, s quarks and antiquarks, keeping the 1981 parameters and comparing with present experimental data.
We show a natural relation between the monodromy formula for focus-focus singularities of integrable Hamiltonian systems and a formula of Duistermaat-Heckman, and extend the main results of our previous note on focus-focus singularities ($\bbS^1$-action, monodromy, and topological classification) to the degenerate case. We also consider the non-Hamiltonian case, local normal forms, etc.
Rotation appears as a dominant effect in massive star evolution. It largely affects all the model outputs: inner structure, tracks, lifetimes, isochrones, surface compositions, blue to red supergiant ratios, etc. At lower metallicities, the effects of rotational mixing are larger; also, more stars may reach critical velocity, even if the initial distribution of rotational velocities is the same.