Search

The potential of deep neural networks in skin lesion classification has already been demonstrated to be on-par if not superior to the dermatologists diagnosis. However, the performance of these models usually deteriorates when the test data differs significantly from the training data (i.e. domain shift). This concerning limitation for models intended to be used in real-world skin lesion classification tasks poses a risk to patients. For example, different image acquisition systems or previously unseen anatomical sites on the patient can suffice to cause such domain shifts. Mitigating the negative effect of such shifts is therefore crucial, but developing effective methods to address domain shift has proven to be challenging. In this study, we carry out an in-depth analysis of eight different unsupervised domain adaptation methods to analyze their effectiveness in improving generalization for dermoscopic datasets. To ensure robustness of our findings, we test each method on a total of ten distinct datasets, thereby covering a variety of possible domain shifts. In addition, we investigated which factors in the domain shifted datasets have an impact on the effectiveness of domain adaptation methods. Our findings show that all of the eight domain adaptation methods result in improved AUPRC for the majority of analyzed datasets. Altogether, these results indicate that unsupervised domain adaptations generally lead to performance improvements for the binary melanoma-nevus classification task regardless of the nature of the domain shift. However, small or heavily imbalanced datasets lead to a reduced conformity of the results due to the influence of these factors on the methods performance.

The limited ability of Convolutional Neural Networks to generalize to images from previously unseen domains is a major limitation, in particular, for safety-critical clinical tasks such as dermoscopic skin cancer classification. In order to translate CNN-based applications into the clinic, it is essential that they are able to adapt to domain shifts. Such new conditions can arise through the use of different image acquisition systems or varying lighting conditions. In dermoscopy, shifts can also occur as a change in patient age or occurence of rare lesion localizations (e.g. palms). These are not prominently represented in most training datasets and can therefore lead to a decrease in performance. In order to verify the generalizability of classification models in real world clinical settings it is crucial to have access to data which mimics such domain shifts. To our knowledge no dermoscopic image dataset exists where such domain shifts are properly described and quantified. We therefore grouped publicly available images from ISIC archive based on their metadata (e.g. acquisition location, lesion localization, patient age) to generate meaningful domains. To verify that these domains are in fact distinct, we used multiple quantification measures to estimate the presence and intensity of domain shifts. Additionally, we analyzed the performance on these domains with and without an unsupervised domain adaptation technique. We observed that in most of our grouped domains, domain shifts in fact exist. Based on our results, we believe these datasets to be helpful for testing the generalization capabilities of dermoscopic skin cancer classifiers.

Document summarization aims to create a precise and coherent summary of a text document. Many deep learning summarization models are developed mainly for English, often requiring a large training corpus and efficient pre-trained language models and tools. However, English summarization models for low-resource Indian languages are often limited by rich morphological variation, syntax, and semantic differences. In this paper, we propose GAE-ISumm, an unsupervised Indic summarization model that extracts summaries from text documents. In particular, our proposed model, GAE-ISumm uses Graph Autoencoder (GAE) to learn text representations and a document summary jointly. We also provide a manually-annotated Telugu summarization dataset TELSUM, to experiment with our model GAE-ISumm. Further, we experiment with the most publicly available Indian language summarization datasets to investigate the effectiveness of GAE-ISumm on other Indian languages. Our experiments of GAE-ISumm in seven languages make the following observations: (i) it is competitive or better than state-of-the-art results on all datasets, (ii) it reports benchmark results on TELSUM, and (iii) the inclusion of positional and cluster information in the proposed model improved the performance of summaries., Comment: 9 pages, 7 figures

Although generative adversarial networks (GANs) have shown promise in medical imaging, they have four main limitations that impeded their utility: computational cost, data requirements, reliable evaluation measures, and training complexity. Our work investigates each of these obstacles in a novel application of StyleGAN2-ADA to high-resolution medical imaging datasets. Our dataset is comprised of liver-containing axial slices from non-contrast and contrast-enhanced computed tomography (CT) scans. Additionally, we utilized four public datasets composed of various imaging modalities. We trained a StyleGAN2 network with transfer learning (from the Flickr-Faces-HQ dataset) and data augmentation (horizontal flipping and adaptive discriminator augmentation). The network's generative quality was measured quantitatively with the Fr\'echet Inception Distance (FID) and qualitatively with a visual Turing test given to seven radiologists and radiation oncologists. The StyleGAN2-ADA network achieved a FID of 5.22 ($\pm$ 0.17) on our liver CT dataset. It also set new record FIDs of 10.78, 3.52, 21.17, and 5.39 on the publicly available SLIVER07, ChestX-ray14, ACDC, and Medical Segmentation Decathlon (brain tumors) datasets. In the visual Turing test, the clinicians rated generated images as real 42% of the time, approaching random guessing. Our computational ablation study revealed that transfer learning and data augmentation stabilize training and improve the perceptual quality of the generated images. We observed the FID to be consistent with human perceptual evaluation of medical images. Finally, our work found that StyleGAN2-ADA consistently produces high-quality results without hyperparameter searches or retraining., Comment: This preprint has not undergone post-submission improvements or corrections. The Version of Record of this contribution is published in LNCS, volume 13570, and is available online at https://doi.org/10.1007/978-3-031-16980-9_14

Graph Convolutional Networks (GCN) have achieved state-of-art results on single text classification tasks like sentiment analysis, emotion detection, etc. However, the performance is achieved by testing and reporting on resource-rich languages like English. Applying GCN for multi-task text classification is an unexplored area. Moreover, training a GCN or adopting an English GCN for Indian languages is often limited by data availability, rich morphological variation, syntax, and semantic differences. In this paper, we study the use of GCN for the Telugu language in single and multi-task settings for four natural language processing (NLP) tasks, viz. sentiment analysis (SA), emotion identification (EI), hate-speech (HS), and sarcasm detection (SAR). In order to evaluate the performance of GCN with one of the Indian languages, Telugu, we analyze the GCN based models with extensive experiments on four downstream tasks. In addition, we created an annotated Telugu dataset, TEL-NLP, for the four NLP tasks. Further, we propose a supervised graph reconstruction method, Multi-Task Text GCN (MT-Text GCN) on the Telugu that leverages to simultaneously (i) learn the low-dimensional word and sentence graph embeddings from word-sentence graph reconstruction using graph autoencoder (GAE) and (ii) perform multi-task text classification using these latent sentence graph embeddings. We argue that our proposed MT-Text GCN achieves significant improvements on TEL-NLP over existing Telugu pretrained word embeddings, and multilingual pretrained Transformer models: mBERT, and XLM-R. On TEL-NLP, we achieve a high F1-score for four NLP tasks: SA (0.84), EI (0.55), HS (0.83) and SAR (0.66). Finally, we show our model's quantitative and qualitative analysis on the four NLP tasks in Telugu., Comment: 9 pages, 6 figures

We report here the discovery of a hot Jupiter at an orbital period of $3.208666\pm0.000016$ days around TOI-1789 (TYC 1962-00303-1, $TESS_{mag}$ = 9.1) based on the TESS photometry, ground-based photometry, and high-precision radial velocity observations. The high-precision radial velocity observations were obtained from the high-resolution spectrographs, PARAS at Physical Research Laboratory (PRL), India, and TCES at Th\"uringer Landessternwarte Tautenburg (TLS), Germany, and the ground-based transit observations were obtained using the 0.43~m telescope at PRL with the Bessel-$R$ filter. The host star is a slightly evolved ($\log{g_*}$ = $3.939^{+0.024}_{-0.046}$), late F-type ($T_{eff}$ = $5984^{+55}_{-57}$ K), metal-rich star ([Fe/H] = $0.370^{+0.073}_{-0.089}$ dex) with a radius of {\ensuremath{$R_{*}$}} = $2.172^{+0.037}_{-0.035}$ \(R_\odot\) located at a distance of $223.56^{+0.91}_{-0.90}$ pc. The simultaneous fitting of the multiple light curves and the radial velocity data of TOI-1789 reveals that TOI-1789b has a mass of $M_{P}$ = $0.70\pm0.16 $ $M_{J}$, a radius of $R_{P}$ = $1.40^{+0.22}_{-0.13}$ $R_{J}$, and a bulk density of $\rho_P$ = $0.31^{+0.15}_{-0.13}$ g cm$^{-3}$ with an orbital separation of a = $0.04873^{+0.00065}_{-0.0016}$ AU. This puts TOI-1789b in the category of inflated hot Jupiters. It is one of the few nearby evolved stars with a close-in planet. The detection of such systems will contribute to our understanding of mechanisms responsible for inflation in hot Jupiters and also provide an opportunity to understand the evolution of planets around stars leaving the main sequence branch., Comment: 15 pages, 12 figures, Accepted for publication in MNRAS

The Echelle spectrograph operating at Vainu Bappu Telescope (VBT), India, is a general purpose instrument used for many high-resolution spectroscopic observations. A concerted effort is being made to expand the scientific capability of the instrument in emerging areas of observational astronomy. The present study is aimed at evaluating the feasibility of the spectrograph to carry out precision Radial Velocity (RV) measurements. In the current design, major factors limiting the RV precision of the spectrograph arise from the movable grating and slit; optical aberrations; positional uncertainty associated with optomechanical mounts and environmental and thermal instabilities in the spectrograph room. RV instabilities due to temperature and pressure variations in the environment are estimated to vary around 120 $ \textrm{ms}^{-1} $ and 400 $ \textrm{ms}^{-1} $ respectively. The positional uncertainty of the grating in the spectrograph could induce a spectral shift $\sim1.4\,\textrm{kms}^{-1} $ across the Echelle orders. A Zemax model is used to overcome the uncertainty in the zero-positioning and lack of repeatability of the moving components. We propose to obtain the ThAr lamp observations and using the Zemax model as the reference, predict the drifts in the positions of the optical components. The perturbations of the optical components from the nominal position are corrected at the beginning of the observational run. After a good match is obtained between the model and the observations, we propose to use a Zemax model to improve the wavelength calibration solution. We could match the observations and model within $\pm$~1 pixels accuracy after the model parameters were perturbed in a real-time setup of the spectrograph. In this paper, we present the estimation of the perturbations of optical components and the effect on the RV obtained.

The Echelle spectrograph operating at Vainu Bappu Telescope (VBT) is a general purpose instrument designed for high resolution spectroscopy. It is being considered for precision Doppler measurements without altering the existing design and basic usage. However, the design level limitations and environmental perturbations are a major source of instability and systematic errors. As a result, a small Doppler signal in the stellar spectra is completely swamped by the large and uncontrolled instrumental drift. In this paper, we discuss some of the remedial measures we took to improve the radial velocity performance of the spectrograph. We show that a new auto-guider assembly has greatly reduced the mechanical jitter of the star image at the fibre input, making the illumination of the spectrograph slit at the other end stable. We have also installed an iodine absorption cell to track and eliminate the instrumental drifts to facilitate precision radial velocity observations. Furthermore, we have developed a generic algorithm that uses iodine exposures to extract the stellar radial velocities without the need for the complex forward modeling. Our algorithm is not accurate to the level of traditional iodine technique. However it is convenient to use on a low-cost general-purpose spectrograph targeting a moderate Radial Velocity (RV) precision at a few 10-100~$\textrm{ms}^{-1}$ level. Finally, we have demonstrated the usefulness of our approach by measuring the RV signal of a well known short-period, planet-hosting star.