arXiv subjects

Siddhant Jain

Publications and source records attributed to Siddhant Jain.

11 recordsLinked to original sources

arXiv2026

Sentiment Analysis of German Sign Language Fairy Tales

We present a dataset and a model for sentiment analysis of German sign language (DGS) fairy tales. First, we perform sentiment analysis for three levels of valence (negative, neutral, positive) on German fairy tales text segments using four large language models (LLMs) and majority voting, reaching an inter-annotator agreement of 0.781 Krippendorff's alpha. Second, we extract face and body motion features from each corresponding DGS video segment using MediaPipe. Finally, we train an explainable model (based on XGBoost) to predict negative, neutral or positive sentiment from video features. Results show an average balanced accuracy of 0.631. A thorough analysis of the most important features reveal that, in addition to eyebrows and mouth motion on the face, also the motion of hips, elbows, and shoulders considerably contribute in the discrimination of the conveyed sentiment, indicating an equal importance of face and body for sentiment communication in sign language.

cs.CL↗

arXiv2026

Convective Preheating Enhances Front Propagation in DCPD Frontal Polymerization

Frontal polymerization (FP) enables rapid curing of thermosets via a self-sustaining thermal wave, but its propagation mechanism can shift dramatically depending on processing conditions. In this study, we investigate the effect of trigger direction and monomer viscosity - controlled via hold time - on the front velocity in frontal ring-opening metathesis polymerization (FROMP) of dicyclopentadiene (DCPD). Our experiments reveal that at low viscosities, bottom-triggered FP fronts propagate significantly faster, ~50% faster front speed compared to top-triggered ones, driven by buoyancy-enhanced convection that preheats the unreacted monomer ahead of the front, that can have important implications for manufacturing applications. However, with increasing hold time, the monomer viscosity rises steeply, suppressing convection and causing the front velocity for top and bottom triggering to converge. This behavior reflects a convection-to-conduction (thermal-diffusion) transition in heat transport during FP. Complementary simulations incorporating buoyancy-driven advection reproduce the observed trends and highlight the importance of fluid flow in front dynamics. These results provide new insight into the coupled thermo-fluid-chemical mechanisms in FP offer strategies to tailor front behavior through viscosity and initiation geometry.

cond-mat.soft↗

arXiv2025

Vortex Dynamics During Pinch-off of Micro-Droplets

Micro-droplets are extensively used in chemical, biological, and medical research, primarily for conducting various tests on samples, including living organisms, using a microfluidic framework. Recent studies have shown that the physiology of bacteria can be significantly altered when subjected to shear and/or extensional stresses. With this motivation, we perform experiments to understand the vortex dynamics involved during the pinch-off process in a cross flow droplet generator, using particle image velocimetry (PIV) to visualize the vortical structures and to quantitatively measure the associated stresses developed inside droplets. The process of pinching off inherently leads to bi-directional acceleration of fluid in the rapidly thinning capillary bridge, resulting in a vortex in the separated droplet as well as in the retracting ligament. We propose scaling laws for the vortical flow inside the droplet post pinch-off and predict the maximum circulation production inside droplet. Further, we discuss the vortex dynamics inside the droplet, the retracting ligament and the advancing ligament and examine the stress fields associated with this transient phenomenon.

physics.flu-dyn↗

arXiv2024

Interaction of vortex ring with perforated plate at different included angles

Experiments are performed to investigate the interaction of a vortex ring (Reynolds number based on circulation (Re_Gamma = 10500) with perforated surface (open area ratio, phi_1 = 0.24 and phi_2 = 0.44) with different included angles (theta = 60deg-180deg). The phenomenon is characterized using techniques like planer laser-induced fluorescence (PLIF) imaging and particle image velocimetry (PIV). Lagrangian analysis using finite-time Lyapunov exponents (FTLE) and Gamma_2 vortex identification methods are utilized to understand flow physics. We observe the development of mushroom-like structures at the holes, driven by the induced flow of the vortex ring. These formations, together with Kelvin-Helmholtz instability, introduce initial instability to the emerging jets. We discern a sequential emergence of the vortex ring in the form of jets at lower theta value that diminishes at higher values. Notably, a single vortex ring is split into two distinct vortex rings for theta less than equal to 120deg. On either side of the perforated plate, the sense of circulation after interaction does not show bias towards the sense of flow on the upstream region for some values of theta. We further show the evolution of circulation by jets in the downstream region aligns with the proposed cumulative slug flow model using the centre line peak velocity of individual jets.

physics.flu-dyn↗

arXiv2024

Video Interpolation with Diffusion Models

We present VIDIM, a generative model for video interpolation, which creates short videos given a start and end frame. In order to achieve high fidelity and generate motions unseen in the input data, VIDIM uses cascaded diffusion models to first generate the target video at low resolution, and then generate the high-resolution video conditioned on the low-resolution generated video. We compare VIDIM to previous state-of-the-art methods on video interpolation, and demonstrate how such works fail in most settings where the underlying motion is complex, nonlinear, or ambiguous while VIDIM can easily handle such cases. We additionally demonstrate how classifier-free guidance on the start and end frame and conditioning the super-resolution model on the original high-resolution frames without additional parameters unlocks high-fidelity results. VIDIM is fast to sample from as it jointly denoises all the frames to be generated, requires less than a billion parameters per diffusion model to produce compelling results, and still enjoys scalability and improved quality at larger parameter counts.

cs.CV↗

arXiv2024

Insights into the mechanics of pure and bacteria-laden sessile whole blood droplet evaporation

We study the mechanics of evaporation and precipitate formation in pure and bacteria-laden sessile whole blood droplets in the context of disease diagnostics. Using experimental and theoretical analysis, we show evaporation process has three stages based on evaporation rate. In the first stage, edge evaporation results in a gelated contact line along the periphery through sol-gel phase transition. The intermediate stage consists of gelated front propagating radially inwards due to capillary flow and droplet height regression in pinned mode, forming a wet-gel phase. We unearthed that the gelation of the entire droplet occurs in the second stage, and the wet-gel formed contains trace amount of water. In the final slowest stage, wet-gel transforms into dry-gel, leading to desiccation-induced stress forming diverse crack patterns in the precipitate. Slow evaporation in the final stage is quantitatively measured using evaporation of trace water and associated transient delamination of the precipitate. Using axisymmetric lubrication approximation, we compute the transient droplet height profile and the erythrocytes concentration for the first two stages of evaporation. We show that the precipitate thickness profile computed from the theoretical analysis conforms to the optical profilometry measurements. We show that the drop evaporation rate and final dried residue pattern do not change appreciably within the parameter variation of the bacterial concentration typically found in bacterial infection of living organisms. However, at exceedingly high bacterial concentrations, the cracks formed in the coronal region deviate from the typical radial cracks found in lower concentrations.

physics.flu-dyn↗

arXiv2024

Dynamics of Soap Bubble Inflation

Bubbles have always captivated our curiosity with their aesthetics and complexities alike. While the act of blowing bubbles is familiar to everyone, the underlying physics of these fleeting spheres often eludes reasoning. In this letter, we discuss the dynamics of inflating a soap bubble using controlled airflow through a film-coated nozzle. We assess and predict the rate of inflation by varying the source pressure. Visualising the previously unexplored internal flow reveals that air enters the bubble as a round jet, emerging from the nozzle opening and impinges on the expanding concave bubble interface to form a toroidal vortex. Several scaling laws of the associated vortical flow spanning the entire bubble and the vortex core are reported. The observed dynamics of this bubble-confined vortex ring formation indicate universality in certain aspects when compared to the free laminar vortex rings.

physics.flu-dyn↗

arXiv2020

Performance evaluation of deep neural networks for forecasting time-series with multiple structural breaks and high volatility

The problem of automatic and accurate forecasting of time-series data has always been an interesting challenge for the machine learning and forecasting community. A majority of the real-world time-series problems have non-stationary characteristics that make the understanding of trend and seasonality difficult. Our interest in this paper is to study the applicability of the popular deep neural networks (DNN) as function approximators for non-stationary TSF. We evaluate the following DNN models: Multi-layer Perceptron (MLP), Convolutional Neural Network (CNN), and RNN with Long-Short Term Memory (LSTM-RNN) and RNN with Gated-Recurrent Unit (GRU-RNN). These DNN methods have been evaluated over 10 popular Indian financial stocks data. Further, the performance evaluation of these DNNs has been carried out in multiple independent runs for two settings of forecasting: (1) single-step forecasting, and (2) multi-step forecasting. These DNN methods show convincing performance for single-step forecasting (one-day ahead forecast). For the multi-step forecasting (multiple days ahead forecast), we have evaluated the methods for different forecast periods. The performance of these methods demonstrates that long forecast periods have an adverse effect on performance.

cs.NE↗

arXiv2019

Few-Shot Point Cloud Region Annotation with Human in the Loop

We propose a point cloud annotation framework that employs human-in-loop learning to enable the creation of large point cloud datasets with per-point annotations. Sparse labels from a human annotator are iteratively propagated to generate a full segmentation of the network by fine-tuning a pre-trained model of an allied task via a few-shot learning paradigm. We show that the proposed framework significantly reduces the amount of human interaction needed in annotating point clouds, without sacrificing on the quality of the annotations. Our experiments also suggest the suitability of the framework in annotating large datasets by noting a reduction in human interaction as the number of full annotations completed by the system increases. Finally, we demonstrate the flexibility of the framework to support multiple different annotations of the same point cloud enabling the creation of datasets with different granularities of annotation.

cs.CV↗

arXiv2018

An Amalgamation of Classical and Quantum Machine Learning For the Classification of Adenocarcinoma and Squamous Cell Carcinoma Patients

The ability to accurately classify disease subtypes is of vital importance, especially in oncology where this capability could have a life saving impact. Here we report a classification between two subtypes of non-small cell lung cancer, namely Adeno- carcinoma vs Squamous cell carcinoma. The data consists of approximately 20,000 gene expression values for each of 104 patients. The data was curated from [1] [2]. We used an amalgamation of classical and and quantum machine learning models to successfully classify these patients. We utilized feature selection methods based on univariate statistics in addition to XGBoost [3]. A novel and proprietary data representation method developed by one of the authors called QCrush was also used as it was designed to incorporate a maximal amount of information under the size constraints of the D-Wave quantum annealing computer. The machine learning was performed by a Quantum Boltzmann Machine. This paper will report our results, the various classical methods, and the quantum machine learning approach we utilized.

stat.ML↗

arXiv2017

BigHand2.2M Benchmark: Hand Pose Dataset and State of the Art Analysis

In this paper we introduce a large-scale hand pose dataset, collected using a novel capture method. Existing datasets are either generated synthetically or captured using depth sensors: synthetic datasets exhibit a certain level of appearance difference from real depth images, and real datasets are limited in quantity and coverage, mainly due to the difficulty to annotate them. We propose a tracking system with six 6D magnetic sensors and inverse kinematics to automatically obtain 21-joints hand pose annotations of depth maps captured with minimal restriction on the range of motion. The capture protocol aims to fully cover the natural hand pose space. As shown in embedding plots, the new dataset exhibits a significantly wider and denser range of hand poses compared to existing benchmarks. Current state-of-the-art methods are evaluated on the dataset, and we demonstrate significant improvements in cross-benchmark performance. We also show significant improvements in egocentric hand pose estimation with a CNN trained on the new dataset.

cs.CV↗