Showing papers by "Sanjay Ghosh published in 2020"

Global Road Damage Detection: State-of-the-art Solutions

Abstract: This paper summarizes the Global Road Damage Detection Challenge (GRDDC), a Big Data Cup organized as a part of the IEEE International Conference on Big Data’2020. The Big Data Cup challenges involve a released dataset and a well-defined problem with clear evaluation metrics. The challenges run on a data competition platform that maintains a leaderboard for the participants. In the presented case, the data constitute 26336 road images collected from India, Japan, and the Czech Republic to propose methods for automatically detecting road damages in these countries. In total, 121 teams from several countries registered for this competition. The submitted solutions were evaluated using two datasets test1 and test2, comprising 2,631 and 2,664 images. This paper encapsulates the top 12 solutions proposed by these teams. The best performing model utilizes YOLO-based ensemble learning to yield an F1 score of 0.67 on test1 and 0.66 on test2. The paper concludes with a review of the facets that worked well for the presented challenge and those that could be improved in future challenges.

Transfer Learning-based Road Damage Detection for Multiple Countries.

Abstract: Many municipalities and road authorities seek to implement automated evaluation of road damage. However, they often lack technology, know-how, and funds to afford state-of-the-art equipment for data collection and analysis of road damages. Although some countries, like Japan, have developed less expensive and readily available Smartphone-based methods for automatic road condition monitoring, other countries still struggle to find efficient solutions. This work makes the following contributions in this context. Firstly, it assesses the usability of the Japanese model for other countries. Secondly, it proposes a large-scale heterogeneous road damage dataset comprising 26620 images collected from multiple countries using smartphones. Thirdly, we propose generalized models capable of detecting and classifying road damages in more than one country. Lastly, we provide recommendations for readers, local agencies, and municipalities of other countries when one other country publishes its data and model for automatic road damage detection and classification. Our dataset is available at (this https URL).

Fast Scale-Adaptive Bilateral Texture Smoothing

Abstract: In the classical bilateral filter, a range kernel is used together with a spatial kernel for smoothing out fine details while simultaneously preserving edges. More recently, it has been demonstrated that even coarse textures can be smoothed using joint bilateral filtering. In this paper, we demonstrate that the superior texture filtering results can be obtained by adapting the spatial kernel at each pixel. To the best of our knowledge, spatial adaptation (of the bilateral filter) has not been explored for texture smoothing. The rationale behind adapting the spatial kernel is that one cannot smooth beyond a certain level using a fixed spatial kernel, no matter how we manipulate the range kernel. In fact, we should simply aggregate more pixels using a sufficiently wide spatial kernel to locally enhance the smoothing. Based on this reasoning, we propose to use the classical bilateral filter for texture smoothing, where we adapt the width of the spatial kernel at each pixel. We describe a simple and efficient gradient-based rule for the latter task. The attractive aspect is that we are able to develop a fast algorithm that can accelerate the computations by an order without visibly compromising the filtering quality. We demonstrate that our method outperforms classical bilateral filtering, joint bilateral filtering, and other filtering methods, and is competitive with the optimization methods. We also present some applications of texture smoothing using the proposed method.

Extraction of built-up areas from Landsat-8 OLI data based on spectral-textural information and feature selection using support vector machine method

Abstract: Information of built-up area is essential for various applications, such as sustainable development or urban planning. Built-up area extraction using optical data is challenging due to spectral con...

Analysis of land use land cover change using a new and existing spectral indices and its impact on normalized land surface temperature

Abstract: In the present study, a new spectral index named as Normalized Ratio Urban Index (NRUIms) has been proposed for multi-sensor Landsat satellite data. The new index along with other existing indices,...

Model-Based Deep Learning for Reconstruction of Joint k-q Under-sampled High Resolution Diffusion MRI

Abstract: We propose a model-based deep learning architecture for the reconstruction of highly accelerated diffusion magnetic resonance imaging (MRI) that enables high resolution imaging. The proposed reconstruction jointly recovers all the diffusion weighted images in a single step from a joint k-q under-sampled acquisition in a parallel MRI setting. We propose the novel use of a pre-trained denoiser as a regularizer in a model-based reconstruction for the recovery of highly under-sampled data. Specifically, we designed the denoiser based on a general diffusion MRI tissue microstructure model for multi-compartmental modeling. By using a wide range of biologically plausible parameter values for the multi-compartmental microstructure model, we simulated diffusion signal that spans the entire microstructure parameter space. A neural network was trained in an unsupervised manner using an autoencoder to learn the diffusion MRI signal subspace. We employed the autoencoder in a model-based reconstruction and show that the autoencoder provides a strong denoising prior to recover the q-space signal. We show reconstruction results on a simulated brain dataset that shows high acceleration capabilities of the proposed method.

Model-Based Deep Learning for Reconstruction of Joint k-q Under-sampled High Resolution Diffusion MRI

Abstract: We propose a model-based deep learning architecture for the reconstruction of highly accelerated diffusion magnetic resonance imaging (MRI) that enables high resolution imaging. The proposed reconstruction jointly recovers all the diffusion weighted images in a single step from a joint k-q under-sampled acquisition in a parallel MRI setting. We propose the novel use of a pre-trained denoiser as a regularizer in a model-based reconstruction for the recovery of highly under-sampled data. Specifically, we designed the denoiser based on a general diffusion MRI tissue microstructure model for multi-compartmental modeling. By using a wide range of biologically plausible parameter values for the multi-compartmental microstructure model, we simulated diffusion signal that spans the entire microstructure parameter space. A neural network was trained in an unsupervised manner using an autoencoder to learn the diffusion MRI signal subspace. We employed the autoencoder in a model-based reconstruction and show that the autoencoder provides a strong denoising prior to recover the q-space signal. We show reconstruction results on a simulated brain dataset that shows high acceleration capabilities of the proposed method.

Liquid-liquid equilibria of the phosphoric acid, ester and water ternary system

Abstract: This paper presents an experimental methodology used for the determination of liquid-liquid equilibrium (LLE) data of the ternary system (phosphoric acid, ester and water). This experimental methodology represents the determination of phosphoric acid linked with different aquatic systems. In this study, an ester was selected as an organic solvent for the recovery of phosphoric acid from waste water. The binodal curve and the tie lines have been prominent. The ternary system (water + phosphoric acid + ester) was studied at three temperatures i.e. 25, 35 and 45 °C (298, 313 and 323 K). The results indicate that the extraction of phosphoric acid by a solvent is possible in aquatic systems. The results are dis- cussed.

Ribosome profiling reveals ribosome stalling on tryptophan codons upon oxidative stress in fission yeast

Abstract: Modulation of translation is an essential response to stress conditions. We have investigated the translational programmes launched by the fission yeast Schizosaccharomyces pombe subject to five environmental stresses: oxidative stress, heavy metal, heat shock, osmotic shock and DNA damage. We also explored the contribution of two major defence pathways to these programmes: The Integrated Stress Response, which directly regulates translation initiation, and the stress-response MAPK pathway. To obtain a genome-wide and high-resolution view of this phenomenon, we performed ribosome profiling of control cells and of cells subject to each of the five stresses mentioned above, both in wild type background and in cells in which the Integrated Stress Response or the MAPK pathway were inactivated. Translational changes were partially dependent on the integrity of both signalling pathways. Interestingly, we found that the transcription factor Fil1, a functional homologue of the Gcn4 and Atf4 proteins (from budding yeast and mammals, respectively), was highly upregulated in most stresses. Consistent with this result, Fil1 was required for the normal response to most stresses. A large group of mRNAs were translationally downregulated, including many required for ribosome biogenesis. Overall, our data suggest that severe stresses lead to the implementation of a universal translational response, which includes energy-saving measures (reduction of ribosome production) and induction of a Fil1-mediated transcriptional programme. Surprisingly, ribosomes stalled on tryptophan codons specifically upon oxidative stress, a phenomenon that is likely caused by a decrease in charged tRNA-Tryptophan. Tryptophan stalling led to a mild translation elongation reduction and contributed to the inhibition of initiation by the Integrated Stress Response. Taken together, our results show that different stresses elicit common and specific translational responses, revealing a special and so far unknown role in Tryptophan-tRNA availability.

Whole Genome Sequences of Aedes aegypti (Linn.) Field Isolates from Southern India

Abstract: Aedes spp. mosquitoes are a major health concern as they transmit several viral pathogens resulting in millions of deaths annually around the world. This is compounded by the emergence of insecticide-resistant strains and global warming, which could expose more than half of the world9s population to Aedes-borne diseases in the future. Therefore, a comprehensive understanding of vector biology and the genomic basis of phenotypes such as insecticide resistance in natural populations are of paramount importance. Here, we sequenced the genome of Aedes aegypti mosquitos sampled from dengue-endemic areas and investigated the genetic variations between the previously reported laboratory-reared strain and our field isolates. The mosquito genomic DNA was used for paired-end sequencing using the Illumina platform. The reads were used for template-based assembly and mapped to the Aedes aegypti reference genome. Stringent parameters and multiple variant calling methods were used to identify unique single nucleotide variants (SNVs) and insertions-deletions (indels) and mapped to the Aedes chromosomes to create a draft consensus genome. Gene Ontology analyses was performed on the variant-enriched genes while two gene families involved in insecticide resistance were used for comparative sequence and phylogenetic analyses. Comparative sequence variant analyses showed that the majority of the high-quality variants in our samples mapped to non-coding regions of the genome, while gene ontology analyses of genic variants revealed enrichment of terms relevant to drug binding and insecticide resistance. Importantly, one mutation implicated in pyrethroid resistance was found in one Aedes sample. This is the first report of genome sequences of A. aegypti field isolates from India which reveals variants specific to the wild population. This is a useful resource which will facilitate development of robust integrated vector control strategies for management of Aedes-borne diseases through genetic manipulation of local mosquito populations.

Nonlocal Co-occurrence for Image Downscaling.

Abstract: Image downscaling is one of the widely used operations in image processing and computer graphics. It was recently demonstrated in the literature that kernel-based convolutional filters could be modified to develop efficient image downscaling algorithms. In this work, we present a new downscaling technique which is based on kernel-based image filtering concept. We propose to use pairwise co-occurrence similarity of the pixelpairs as the range kernel similarity in the filtering operation. The co-occurrence of the pixel-pair is learned directly from the input image. This co-occurrence learning is performed in a neighborhood based fashion all over the image. The proposed method can preserve the high-frequency structures, which were present in the input image, into the downscaled image. The resulting images retain visually important details and do not suffer from edge-blurring artifact. We demonstrate the effectiveness of our proposed approach with extensive experiments on a large number of images downscaled with various downscaling factors.