Search

As machine learning gets deployed more and more widely, and model sizes continue to grow, improving computational efficiency during model inference has become a key challenge. In many commonly used model architectures, including Transformers, a significant portion of the inference computation is comprised of exponential non-linearities such as Softmax. In this work, we develop QuAKE, a collection of novel operators that leverage certain properties of IEEE-754 floating point representations to quickly approximate the exponential function without requiring specialized hardware, extra memory, or precomputation. We propose optimizations that enhance the efficiency of QuAKE in commonly used exponential non-linearities such as Softmax, GELU, and the Logistic function. Our benchmarks demonstrate substantial inference speed improvements between 10% and 35% on server CPUs, and 5% and 45% on embedded and mobile-scale CPUs for a variety of model architectures and sizes. Evaluations of model performance on standard datasets and tasks from various domains show that QuAKE operators are able to provide sizable speed benefits with little to no loss of performance on downstream tasks.

Video foreground segmentation (VFS) is an important computer vision task wherein one aims to segment the objects under motion from the background. Most of the current methods are image-based, i.e., rely only on spatial cues while ignoring motion cues. Therefore, they tend to overfit the training data and don't generalize well to out-of-domain (OOD) distribution. To solve the above problem, prior works exploited several cues such as optical flow, background subtraction mask, etc. However, having a video data with annotations like optical flow is a challenging task. In this paper, we utilize the temporal information and the spatial cues from the video data to improve OOD performance. However, the challenge lies in how we model the temporal information given the video data in an interpretable way creates a very noticeable difference. We therefore devise a strategy that integrates the temporal context of the video in the development of VFS. Our approach give rise to deep learning architectures, namely MUSTAN1 and MUSTAN2 and they are based on the idea of multi-scale temporal context as an attention, i.e., aids our models to learn better representations that are beneficial for VFS. Further, we introduce a new video dataset, namely Indoor Surveillance Dataset (ISD) for VFS. It has multiple annotations on a frame level such as foreground binary mask, depth map, and instance semantic annotations. Therefore, ISD can benefit other computer vision tasks. We validate the efficacy of our architectures and compare the performance with baselines. We demonstrate that proposed methods significantly outperform the benchmark methods on OOD. In addition, the performance of MUSTAN2 is significantly improved on certain video categories on OOD data due to ISD., Comment: 10 pages, 8 figures

Rydberg excitons (analogues of Rydberg atoms in condensed matter systems) are highly excited bound electron-hole states with large Bohr radii. The interaction between them as well as exciton coupling to light may lead to strong optical nonlinearity, with applications in sensing and quantum information processing. Here, we achieve strong effective photon-photon interactions (Kerr-like optical nonlinearity) via the Rydberg blockade phenomenon and the hybridisation of excitons and photons forming polaritons in a Cu$_2$O-filled microresonators. Under pulsed resonant excitation polariton resonance frequencies are renormalised due to the reduction of the photon-exciton coupling with increasing exciton density. Theoretical analysis shows that the Rydberg blockade plays a major role in the experimentally observed scaling of the polariton nonlinearity coefficient as $\propto n^{4.4 \pm 1.8}$ for principal quantum numbers up to n = 7. Such high principal quantum numbers studied in a polariton system for the first time are essential for realisation of high Rydberg optical nonlinearities, which paves the way towards quantum optical applications and fundamental studies of strongly-correlated photonic (polaritonic) states in a solid state system.

Audio Question Answering (AQA) constitutes a pivotal task in which machines analyze both audio signals and natural language questions to produce precise natural language answers. The significance of possessing high-quality, diverse, and extensive AQA datasets cannot be overstated when aiming for the precision of an AQA system. While there has been notable focus on developing accurate and efficient AQA models, the creation of high-quality, diverse, and extensive datasets for the specific task at hand has not garnered considerable attention. To address this challenge, this work makes several contributions. We introduce a scalable AQA data generation pipeline, denoted as the AQUALLM framework, which relies on Large Language Models (LLMs). This framework utilizes existing audio-caption annotations and incorporates state-of-the-art LLMs to generate expansive, high-quality AQA datasets. Additionally, we present three extensive and high-quality benchmark datasets for AQA, contributing significantly to the progression of AQA research. AQA models trained on the proposed datasets set superior benchmarks compared to the existing state-of-the-art. Moreover, models trained on our datasets demonstrate enhanced generalizability when compared to models trained using human-annotated AQA data. Code and datasets will be accessible on GitHub~\footnote{\url{https://github.com/swarupbehera/AQUALLM}}.

The task of Prior Case Retrieval (PCR) in the legal domain is about automatically citing relevant (based on facts and precedence) prior legal cases in a given query case. To further promote research in PCR, in this paper, we propose a new large benchmark (in English) for the PCR task: IL-PCR (Indian Legal Prior Case Retrieval) corpus. Given the complex nature of case relevance and the long size of legal documents, BM25 remains a strong baseline for ranking the cited prior documents. In this work, we explore the role of events in legal case retrieval and propose an unsupervised retrieval method-based pipeline U-CREAT (Unsupervised Case Retrieval using Events Extraction). We find that the proposed unsupervised retrieval method significantly increases performance compared to BM25 and makes retrieval faster by a considerable margin, making it applicable to real-time case retrieval systems. Our proposed system is generic, we show that it generalizes across two different legal systems (Indian and Canadian), and it shows state-of-the-art performance on the benchmarks for both the legal systems (IL-PCR and COLIEE corpora)., Comment: Accepted at ACL 2023, 15 pages (12 main + 3 Appendix)

In populous countries, pending legal cases have been growing exponentially. There is a need for developing NLP-based techniques for processing and automatically understanding legal documents. To promote research in the area of Legal NLP we organized the shared task LegalEval - Understanding Legal Texts at SemEval 2023. LegalEval task has three sub-tasks: Task-A (Rhetorical Roles Labeling) is about automatically structuring legal documents into semantically coherent units, Task-B (Legal Named Entity Recognition) deals with identifying relevant entities in a legal document and Task-C (Court Judgement Prediction with Explanation) explores the possibility of automatically predicting the outcome of a legal case along with providing an explanation for the prediction. In total 26 teams (approx. 100 participants spread across the world) submitted systems paper. In each of the sub-tasks, the proposed systems outperformed the baselines; however, there is a lot of scope for improvement. This paper describes the tasks, and analyzes techniques proposed by various teams., Comment: 13 Pages (9 Pages + References), Accepted at SemEval 2023 at ACL 2023

Computational Paninian Grammar model helps in decoding a natural language expression as a series of modifier-modified relations and therefore facilitates in identifying dependency relations closer to language (context) semantics compared to the usual Stanford dependency relations. However, the importance of this CPG dependency scheme has not been studied in the context of multi-modal vision and language applications. At IIIT Hyderabad, we plan to perform a novel study to explore the potential advantages and disadvantages of CPG framework in a vision-language navigation task setting, a popular and challenging multi-modal grounding task.

Treatment decisions for brain metastatic disease rely on knowledge of the primary organ site, and currently made with biopsy and histology. Here we develop a novel deep learning approach for accurate non-invasive digital histology with whole-brain MRI data. Our IRB-approved single-site retrospective study was comprised of patients (n=1,399) referred for MRI treatment-planning and gamma knife radiosurgery over 21 years. Contrast-enhanced T1-weighted and T2-weighted Fluid-Attenuated Inversion Recovery brain MRI exams (n=1,582) were preprocessed and input to the proposed deep learning workflow for tumor segmentation, modality transfer, and primary site classification into one of five classes. Ten-fold cross-validation generated overall AUC of 0.878 (95%CI:0.873,0.883), lung class AUC of 0.889 (95%CI:0.883,0.895), breast class AUC of 0.873 (95%CI:0.860,0.886), melanoma class AUC of 0.852 (95%CI:0.842,0.862), renal class AUC of 0.830 (95%CI:0.809,0.851), and other class AUC of 0.822 (95%CI:0.805,0.839). These data establish that whole-brain imaging features are discriminative to allow accurate diagnosis of the primary organ site of malignancy. Our end-to-end deep radiomic approach has great potential for classifying metastatic tumor types from whole-brain MRI images. Further refinement may offer an invaluable clinical tool to expedite primary cancer site identification for precision treatment and improved outcomes.

In this paper, we focus on learning product graphs from multi-domain data. We assume that the product graph is formed by the Cartesian product of two smaller graphs, which we refer to as graph factors. We pose the product graph learning problem as the problem of estimating the graph factor Laplacian matrices. To capture local interactions in data, we seek sparse graph factors and assume a smoothness model for data. We propose an efficient iterative solver for learning sparse product graphs from data. We then extend this solver to infer multi-component graph factors with applications to product graph clustering by imposing rank constraints on the graph Laplacian matrices. Although working with smaller graph factors is computationally more attractive, not all graphs may readily admit an exact Cartesian product factorization. To this end, we propose efficient algorithms to approximate a graph by a nearest Cartesian product of two smaller graphs. The efficacy of the developed framework is demonstrated using several numerical experiments on synthetic data and real data., Comment: 13 pages, 5 figures. Submitted to TSP

The quantum confinement of Rydberg excitons is an important step towards exploiting their large nonlinearities for quantum applications. We observe Rydberg excitons in natural nanoparticles of Cu$_2$O. We resolve up to the principal quantum number $n=12$ in a bulk Cu$_2$O crystal and up to $n=6$ in nanoparticles extracted from the same crystal. The exciton transitions in nanoparticles are broadened and their oscillator strengths decrease as $\propto n^{-4}$ compared to those in the bulk (decreasing as $\propto n^{-3}$). We explain our results by including the effect of quantum confinement of exciton states in the nanoparticles. Our results provide an understanding of the physics of Cu$_2$O Rydberg excitons in confined dimensions., Comment: 5 pages, 5 figures

Machine learning is widely used in developing computer-aided diagnosis (CAD) schemes of medical images. However, CAD usually computes large number of image features from the targeted regions, which creates a challenge of how to identify a small and optimal feature vector to build robust machine learning models. In this study, we investigate feasibility of applying a random projection algorithm to build an optimal feature vector from the initially CAD-generated large feature pool and improve performance of machine learning model. We assemble a retrospective dataset involving 1,487 cases of mammograms in which 644 cases have confirmed malignant mass lesions and 843 have benign lesions. A CAD scheme is first applied to segment mass regions and initially compute 181 features. Then, support vector machine (SVM) models embedded with several feature dimensionality reduction methods are built to predict likelihood of lesions being malignant. All SVM models are trained and tested using a leave-one-case-out cross-validation method. SVM generates a likelihood score of each segmented mass region depicting on one-view mammogram. By fusion of two scores of the same mass depicting on two-view mammograms, a case-based likelihood score is also evaluated. Comparing with the principle component analyses, nonnegative matrix factorization, and Chi-squared methods, SVM embedded with the random projection algorithm yielded a significantly higher case-based lesion classification performance with the area under ROC curve of 0.84+0.01 (p<0.02). The study demonstrates that the random project algorithm is a promising method to generate optimal feature vectors to help improve performance of machine learning models of medical images., Comment: 11 pages, 6 figures

This paper proposes a Fast Graph Convolutional Neural Network (FGRNN) architecture to predict sequences with an underlying graph structure. The proposed architecture addresses the limitations of the standard recurrent neural network (RNN), namely, vanishing and exploding gradients, causing numerical instabilities during training. State-of-the-art architectures that combine gated RNN architectures, such as Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) with graph convolutions are known to improve the numerical stability during the training phase, but at the expense of the model size involving a large number of training parameters. FGRNN addresses this problem by adding a weighted residual connection with only two extra training parameters as compared to the standard RNN. Numerical experiments on the real 3D point cloud dataset corroborates the proposed architecture., Comment: 5 pages.Submitted to Asilomar Conference on Signals, Systems, and Computers

In this paper, we focus on learning the underlying product graph structure from multidomain training data. We assume that the product graph is formed from a Cartesian graph product of two smaller factor graphs. We then pose the product graph learning problem as the factor graph Laplacian matrix estimation problem. To estimate the factor graph Laplacian matrices, we assume that the data is smooth with respect to the underlying product graph. When the training data is noise free or complete, learning factor graphs can be formulated as a convex optimization problem, which has an explicit solution based on the water-filling algorithm. The developed framework is illustrated using numerical experiments on synthetic data as well as real data related to air quality monitoring in India., Comment: Submitted to ICASSP 2020

SimRank is a widely studied link-based similarity measure that is known for its simple, yet powerful philosophy that two nodes are similar if they are referenced by similar nodes. While this philosophy has been the basis of several improvements, there is another useful, albeit less frequently discussed interpretation for SimRank known as the Random Surfer-Pair Model. In this work, we show that other well known measures related to SimRank can also be reinterpreted using Random Surfer-Pair Models, and establish a mathematically sound, general and unifying framework for several link-based similarity measures. This also serves to provide new insights into their functioning and allows for using these measures in a Monte Carlo framework, which provides several computational benefits. Next, we describe how the framework can be used as a versatile tool for developing measures according to given design requirements. As an illustration of this utility, we develop a new measure by combining the benefits of two existing measures under this framework, and empirically demonstrate that it results in a better performing measure., Comment: Corrected affiliations

A decade ago, non-radiative wireless power transmission re-emerged as a promising alternative to deliver electrical power to devices where a physical wiring proved to be unfeasible. However, existing approaches are neither scalable nor efficient when multiple devices are involved, as they are restricted by factors like coupling and external environments. Zenneck waves are excited at interfaces, like surface plasmons and have the potential to deliver electrical power to devices placed on a conducting surface. Here, we demonstrate, efficient and long range delivery of electrical power by exciting nonradiative waves over metal surfaces to multiple loads. Our modeling and simulation using Maxwells equation with proper boundary conditions shows Zenneck type behavior for the excited waves and are in excellent agreement with experimental results. In conclusion, we physically realize a radically different power transfer system, based on a wave, whose existence has been fiercely debated for over a century., Comment: 16 pages, 4 figures. Article submitted to Nature Communications

Robust Policy Search is the problem of learning policies that do not degrade in performance when subject to unseen environment model parameters. It is particularly relevant for transferring policies learned in a simulation environment to the real world. Several existing approaches involve sampling large batches of trajectories which reflect the differences in various possible environments, and then selecting some subset of these to learn robust policies, such as the ones that result in the worst performance. We propose an active learning based framework, EffAcTS, to selectively choose model parameters for this purpose so as to collect only as much data as necessary to select such a subset. We apply this framework using Linear Bandits, and experimentally validate the gains in sample efficiency and the performance of our approach on standard continuous control tasks. We also present a Multi-Task Learning perspective to the problem of Robust Policy Search, and draw connections from our proposed framework to existing work on Multi-Task Learning., Comment: 9 pages, 6 figures

This paper presents a method called One-class Classification using Length statistics of Emerging Patterns Plus (OCLEP+)., Comment: 13 pages

Memristors based on a double barrier design have been analysed by various nano spectroscopic methods to unveil details about its microstructure and conduction mechanism. The device consists of an AlOx tunnel barrier and a NbOy/Au Schottky barrier sandwiched between Nb bottom electrode and Au top electrode. As it was anticipated that the local chemical composition of the tunnel barrier, i.e. oxidation state of the metals as well as concentration and distribution of oxygen ions, have a major influence on electronic conduction, these factors were carefully analysed. A combined approach was chosen in order to reliably investigate electronic states of Nb and O by electron energy-loss spectroscopy as well as map elements whose transition edges exhibit a different energy range by energy-dispersive X-ray spectroscopy like Au and Al. The results conclusively demonstrate significant oxidation of the bottom electrode as well as a small oxygen vacancy concentration in the Al oxide tunnel barrier. Possible scenarios to explain this unexpected additional oxide layer are discussed and kinetic Monte Carlo simulations were applied in order to identify its influence on conduction mechanisms in the device. In light of the strong deviations between observed and originally sought layout, this study highlights the robustness in terms of structural deviations of the double barrier memristor device.

We consider an input queued switch operating under the MaxWeight scheduling algorithm. This system is interesting to study because it is a model for Internet routers and data center networks. Recently, it was shown that the MaxWeight algorithm has optimal heavy-traffic queue length scaling when all ports are uniformly saturated. Here we consider the case when an arbitrary number of ports are saturated (which we call the incompletely saturated case), and each port is allowed to saturate at a different rate. We use a recently developed drift technique to show that the heavy-traffic queue length under the MaxWeight scheduling algorithm has optimal scaling with respect to the switch size even in these cases., Comment: 32 pages