arXiv subjects

Nitin Agarwal

Publications and source records attributed to Nitin Agarwal.

25 records · Page 2Linked to original sources

arXiv2020

Image-Driven Furniture Style for Interactive 3D Scene Modeling

Creating realistic styled spaces is a complex task, which involves design know-how for what furniture pieces go well together. Interior style follows abstract rules involving color, geometry and other visual elements. Following such rules, users manually select similar-style items from large repositories of 3D furniture models, a process which is both laborious and time-consuming. We propose a method for fast-tracking style-similarity tasks, by learning a furniture's style-compatibility from interior scene images. Such images contain more style information than images depicting single furniture. To understand style, we train a deep learning network on a classification task. Based on image embeddings extracted from our network, we measure stylistic compatibility of furniture. We demonstrate our method with several 3D model style-compatibility results, and with an interactive system for modeling style-consistent scenes.

cs.CV↗

arXiv2020

GAMesh: Guided and Augmented Meshing for Deep Point Networks

We present a new meshing algorithm called guided and augmented meshing, GAMesh, which uses a mesh prior to generate a surface for the output points of a point network. By projecting the output points onto this prior and simplifying the resulting mesh, GAMesh ensures a surface with the same topology as the mesh prior but whose geometric fidelity is controlled by the point network. This makes GAMesh independent of both the density and distribution of the output points, a common artifact in traditional surface reconstruction algorithms. We show that such a separation of geometry from topology can have several advantages especially in single-view shape prediction, fair evaluation of point networks and reconstructing surfaces for networks which output sparse point clouds. We further show that by training point networks with GAMesh, we can directly optimize the vertex positions to generate adaptive meshes with arbitrary topologies.

cs.CV↗

arXiv2019

Learning Embedding of 3D models with Quadric Loss

Sharp features such as edges and corners play an important role in the perception of 3D models. In order to capture them better, we propose quadric loss, a point-surface loss function, which minimizes the quadric error between the reconstructed points and the input surface. Computation of Quadric loss is easy, efficient since the quadric matrices can be computed apriori, and is fully differentiable, making quadric loss suitable for training point and mesh based architectures. Through extensive experiments we show the merits and demerits of quadric loss. When combined with Chamfer loss, quadric loss achieves better reconstruction results as compared to any one of them or other point-surface loss functions.

cs.CV↗

arXiv2018

Linked Causal Variational Autoencoder for Inferring Paired Spillover Effects

Modeling spillover effects from observational data is an important problem in economics, business, and other fields of research. % It helps us infer the causality between two seemingly unrelated set of events. For example, if consumer spending in the United States declines, it has spillover effects on economies that depend on the U.S. as their largest export market. In this paper, we aim to infer the causation that results in spillover effects between pairs of entities (or units), we call this effect as \textit{paired spillover}. To achieve this, we leverage the recent developments in variational inference and deep learning techniques to propose a generative model called Linked Causal Variational Autoencoder (LCVA). Similar to variational autoencoders (VAE), LCVA incorporates an encoder neural network to learn the latent attributes and a decoder network to reconstruct the inputs. However, unlike VAE, LCVA treats the \textit{latent attributes as confounders that are assumed to affect both the treatment and the outcome of units}. Specifically, given a pair of units $u$ and $\bar{u}$, their individual treatment and outcomes, the encoder network of LCVA samples the confounders by conditioning on the observed covariates of $u$, the treatments of both $u$ and $\bar{u}$ and the outcome of $u$. Once inferred, the latent attributes (or confounders) of $u$ captures the spillover effect of $\bar{u}$ on $u$. Using a network of users from job training dataset (LaLonde (1986)) and co-purchase dataset from Amazon e-commerce domain, we show that LCVA is significantly more robust than existing methods in capturing spillover effects.

cs.LG↗

arXiv2018

Metabolomic signature of type 1 diabetes-induced sensory loss and nerve damage in diabetic neuropathy

Diabetic-induced peripheral neuropathy (DPN) is a diabetic late complication. The molecular mechanisms underlying the pathophysiology of nerve damage & sensory loss remain largely unclear. Recently, alterations in metabolic flux have gained attention a basis for organ damage in diabetes; however, peripheral sensory neurons have not been adequately analyzed. In the present study, we attempted to delineate the role of alteration of metabolic pathways in relation to nerve damage & sensory loss. We employed STZ-injected mouse model of type1 diabetes. To investigate the progression of DPN by behavioral measurements of sensitivity to thermal & mechanical stimuli and quantitative assessment of intraepidermal nerve fiber density. We employed a MS-based screen to address alterations in levels of metabolites in peripheral sciatic nerve (SN) & amino acids (AA) in serum over several months post-STZ administration. Although hyperglycemia & body weight changes occurred early, sensory loss & reduced intraepithelial branching of nociceptive nerves was only evident at 22 wks post-STZ. The longitudinal metabolites screen in SN demonstrated that mice at 12 and 22 wks post-STZ showed an early impairment the tricarboxylic acid. We found that levels of citric acid, ketoglutaric acid, succinic acid, fumaric acid & malic acid were observed to be significantly reduced in SN at 22 wks post-STZ. In addition, we also found the increase in levels of sorbitol & L-Lactate in SN from 12 wks post-STZ injection. AA screen in serum showed that the amino acids Val, Ile and Leu, increased more than 2-fold from 12 wks post-STZ. Similarly, the levels of Tyr, Asn, Ser, His, Ala, & Pro showed progressive increase. Our results indicate that the impaired TCA cycle metabolites in peripheral nerve is the primary cause of shunting metabolic substrate to compensatory pathways which leads to mitochondrial dysfunction & nerve damage.

q-bio.CB↗

arXiv2018

Hypoxia-inducible factor 1a protects peripheral sensory neurons from diabetic peripheral neuropathy by suppressing accumulation of reactive oxygen species

Diabetic peripheral neuropathy (DPN) is one of the most common diabetic complications. Mechanisms underlying nerve damage and sensory loss following metabolic dysfunction remain large unclear. Recently, hyperglycemia-induced mitochondrial dysfunction and the generation of ROS have gained attention as possible mechanisms of organ damage in diabetes. Hypoxia-inducible factor 1(HIF1a) is a key transcription factor activated by hypoxia, hyperglycemia, nitric oxide as well as ROS, suggesting a fundamental role in DPN susceptibility. Genetically-modified mutant mice, which conditionally lack HIF1a in peripheral sensory neurons (SNS-HIF1a-/-), were analyzed longitudinally up to 6 months in the streptozotocin (STZ) model of type1 diabetes. Behavioral measurements of sensitivity to thermal and mechanical stimuli, quantitative morphological analyses of intraepidermal nerve fiber density and measurements of reactive oxygen species (ROS) in sensory neurons in vivo were undertaken over several months post-STZ injections to delineate the role of HIF1a in DPN. Longitudinal behavioral and morphological analyses at 5, 13 and 24 wks post-STZ treatment revealed that SNS-HIF1a-/- developed stronger hyperglycemia-evoked losses of peripheral nociceptive sensory axons associated with stronger losses of mechano- and heat sensation with a faster onset than HIF1afl/fl mice. Mechanistically, these histomorphologic and behavioral differences were associated with significantly higher level of STZ-induced production of ROS in sensory neurons of SNS-HIF1a-/- mice as compared with HIF1afl/fl. Our results indicate that HIF1a is as an upstream modulator of ROS in peripheral sensory neurons and exerts a protective function in suppressing hyperglycemia-induced nerve damage by limiting ROS levels. HIF1a stabilization may be thus a new strategy target for limiting sensory loss, a debilitating late complication of diabetes.

q-bio.CB↗

arXiv2017

Geometry Processing of Conventionally Produced Mouse Brain Slice Images

Brain mapping research in most neuroanatomical laboratories relies on conventional processing techniques, which often introduce histological artifacts such as tissue tears and tissue loss. In this paper we present techniques and algorithms for automatic registration and 3D reconstruction of conventionally produced mouse brain slices in a standardized atlas space. This is achieved first by constructing a virtual 3D mouse brain model from annotated slices of Allen Reference Atlas (ARA). Virtual re-slicing of the reconstructed model generates ARA-based slice images corresponding to the microscopic images of histological brain sections. These image pairs are aligned using a geometric approach through contour images. Histological artifacts in the microscopic images are detected and removed using Constrained Delaunay Triangulation before performing global alignment. Finally, non-linear registration is performed by solving Laplace's equation with Dirichlet boundary conditions. Our methods provide significant improvements over previously reported registration techniques for the tested slices in 3D space, especially on slices with significant histological artifacts. Further, as an application we count the number of neurons in various anatomical regions using a dataset of 51 microscopic slices from a single mouse brain. This work represents a significant contribution to this subfield of neuroscience as it provides tools to neuroanatomist for analyzing and processing histological data.

cs.GR↗