arXiv subjects

Debayan Gupta

Publications and source records attributed to Debayan Gupta.

36 records · Page 2Linked to original sources

arXiv2024

Visual Concept Networks: A Graph-Based Approach to Detecting Anomalous Data in Deep Neural Networks

Deep neural networks (DNNs), while increasingly deployed in many applications, struggle with robustness against anomalous and out-of-distribution (OOD) data. Current OOD benchmarks often oversimplify, focusing on single-object tasks and not fully representing complex real-world anomalies. This paper introduces a new, straightforward method employing graph structures and topological features to effectively detect both far-OOD and near-OOD data. We convert images into networks of interconnected human understandable features or visual concepts. Through extensive testing on two novel tasks, including ablation studies with large vocabularies and diverse tasks, we demonstrate the method's effectiveness. This approach enhances DNN resilience to OOD data and promises improved performance in various applications.

cs.CV↗

arXiv2024

SimSAM: Simple Siamese Representations Based Semantic Affinity Matrix for Unsupervised Image Segmentation

Recent developments in self-supervised learning (SSL) have made it possible to learn data representations without the need for annotations. Inspired by the non-contrastive SSL approach (SimSiam), we introduce a novel framework SIMSAM to compute the Semantic Affinity Matrix, which is significant for unsupervised image segmentation. Given an image, SIMSAM first extracts features using pre-trained DINO-ViT, then projects the features to predict the correlations of dense features in a non-contrastive way. We show applications of the Semantic Affinity Matrix in object segmentation and semantic segmentation tasks. Our code is available at https://github.com/chandagrover/SimSAM.

cs.CV↗

arXiv2024

Generating Synthetic Population

In this paper, we provide a method to generate synthetic population at various administrative levels for a country like India. This synthetic population is created using machine learning and statistical methods applied to survey data such as Census of India 2011, IHDS-II, NSS-68th round, GPW etc. The synthetic population defines individuals in the population with characteristics such as age, gender, height, weight, home and work location, household structure, preexisting health conditions, socio-economical status, and employment. We used the proposed method to generate the synthetic population for various districts of India. We also compare this synthetic population with source data using various metrics. The experiment results show that the synthetic data can realistically simulate the population for various districts of India.

cs.CY↗

arXiv2024

Synthpop++: A Hybrid Framework for Generating A Country-scale Synthetic Population

Population censuses are vital to public policy decision-making. They provide insight into human resources, demography, culture, and economic structure at local, regional, and national levels. However, such surveys are very expensive (especially for low and middle-income countries with high populations, such as India), time-consuming, and may also raise privacy concerns, depending upon the kinds of data collected. In light of these issues, we introduce SynthPop++, a novel hybrid framework, which can combine data from multiple real-world surveys (with different, partially overlapping sets of attributes) to produce a real-scale synthetic population of humans. Critically, our population maintains family structures comprising individuals with demographic, socioeconomic, health, and geolocation attributes: this means that our ``fake'' people live in realistic locations, have realistic families, etc. Such data can be used for a variety of purposes: we explore one such use case, Agent-based modelling of infectious disease in India. To gauge the quality of our synthetic population, we use both machine learning and statistical metrics. Our experimental results show that synthetic population can realistically simulate the population for various administrative units of India, producing real-scale, detailed data at the desired level of zoom -- from cities, to districts, to states, eventually combining to form a country-scale synthetic population.

cs.LG↗

arXiv2023

Developmental Pretraining (DPT) for Image Classification Networks

In the backdrop of increasing data requirements of Deep Neural Networks for object recognition that is growing more untenable by the day, we present Developmental PreTraining (DPT) as a possible solution. DPT is designed as a curriculum-based pre-training approach designed to rival traditional pre-training techniques that are data-hungry. These training approaches also introduce unnecessary features that could be misleading when the network is employed in a downstream classification task where the data is sufficiently different from the pre-training data and is scarce. We design the curriculum for DPT by drawing inspiration from human infant visual development. DPT employs a phased approach where carefully-selected primitive and universal features like edges and shapes are taught to the network participating in our pre-training regime. A model that underwent the DPT regime is tested against models with randomised weights to evaluate the viability of DPT.

cs.LG↗

arXiv2023

RANDGENER: Distributed Randomness Beacon from Verifiable Delay Function

Buoyed by the excitement around secure decentralized applications, the last few decades have seen numerous constructions of distributed randomness beacons (DRB) along with use cases; however, a secure DRB (in many variations) remains an open problem. We further note that it is natural to want some kind of reward for participants who spend time and energy evaluating the randomness beacon value -- this is already common in distributed protocols. In this work, we present RandGener, a novel $n$-party commit-reveal-recover (or collaborative) DRB protocol with a novel reward and penalty mechanism along with a set of realistic guarantees. We design our protocol using trapdoor watermarkable verifiable delay functions in the RSA group setting (without requiring a trusted dealer or distributed key generation).

cs.CR↗

arXiv2023

Trenchcoat: Human-Computable Hashing Algorithms for Password Generation

The average user has between 90-130 online accounts, and around $3 \times 10^{11}$ passwords are in use this year. Most people are terrible at remembering "random" passwords, so they reuse or create similar passwords using a combination of predictable words, numbers, and symbols. Previous password-generation or management protocols have imposed so large a cognitive load that users have abandoned them in favor of insecure yet simpler methods (e.g., writing them down or reusing minor variants). We describe a range of candidate human-computable "hash" functions suitable for use as password generators - as long as the human (with minimal education assumptions) keeps a single, easily-memorizable "master" secret - and rate them by various metrics, including effective security. These functions hash master-secrets with user accounts to produce sub-secrets that can be used as passwords; $F_R($s$, w) \longrightarrow y$, takes a website $w$, produces a password $y$, parameterized by master secret $s$, which may or may not be a string. We exploit the unique configuration $R$ of each user's associative and implicit memory (detailed in section 2) to ensure that sources of randomness unique to each user are present in each master-secret $F_R$. An adversary cannot compute or verify $F_R$ efficiently since $R$ is unique to each individual; in that sense, our hash function is similar to a physically unclonable function. For the algorithms we propose, the user need only complete primitive operations such as addition, spatial navigation or searching. Critically, most of our methods are also accessible to neurodiverse, or cognitively or physically differently-abled persons. We present results from a survey (n=134 individuals) investigating real-world usage of these methods and how people currently come up with their passwords, we also survey 400 websites to collate current password advice.

cs.CR↗

arXiv2023

Tight Short-Lived Signatures

A Time-lock puzzle (TLP) sends information into the future: a predetermined number of sequential computations must occur (i.e., a predetermined amount of time must pass) to retrieve the information, regardless of parallelization. Buoyed by the excitement around secure decentralized applications and cryptocurrencies, the last decade has witnessed numerous constructions of TLP variants and related applications (e.g., cost-efficient blockchain designs, randomness beacons, e-voting, etc.). In this poster, we first extend the notion of TLP by formally defining the "time-lock public key encryption" (TLPKE) scheme. Next, we introduce and construct a "tight short-lived signatures" scheme using our TLPKE. Furthermore, to test the validity of our proposed schemes, we do a proof-of-concept implementation and run detailed simulations.

cs.CR↗

arXiv2023

Sem-CS: Semantic CLIPStyler for Text-Based Image Style Transfer

CLIPStyler demonstrated image style transfer with realistic textures using only a style text description (instead of requiring a reference style image). However, the ground semantics of objects in the style transfer output is lost due to style spill-over on salient and background objects (content mismatch) or over-stylization. To solve this, we propose Semantic CLIPStyler (Sem-CS), that performs semantic style transfer. Sem-CS first segments the content image into salient and non-salient objects and then transfers artistic style based on a given style text description. The semantic style transfer is achieved using global foreground loss (for salient objects) and global background loss (for non-salient objects). Our empirical results, including DISTS, NIMA and user study scores, show that our proposed framework yields superior qualitative and quantitative performance. Our code is available at github.com/chandagrover/sem-cs.

cs.CV↗

arXiv2023

SEM-CS: Semantic CLIPStyler for Text-Based Image Style Transfer

CLIPStyler demonstrated image style transfer with realistic textures using only the style text description (instead of requiring a reference style image). However, the ground semantics of objects in style transfer output is lost due to style spillover on salient and background objects (content mismatch) or over-stylization. To solve this, we propose Semantic CLIPStyler (Sem-CS) that performs semantic style transfer. Sem-CS first segments the content image into salient and non-salient objects and then transfers artistic style based on a given style text description. The semantic style transfer is achieved using global foreground loss (for salient objects) and global background loss (for non-salient objects). Our empirical results, including DISTS, NIMA and user study scores, show that our proposed framework yields superior qualitative and quantitative performance.

cs.CV↗

arXiv2022

ContextCLIP: Contextual Alignment of Image-Text pairs on CLIP visual representations

State-of-the-art empirical work has shown that visual representations learned by deep neural networks are robust in nature and capable of performing classification tasks on diverse datasets. For example, CLIP demonstrated zero-shot transfer performance on multiple datasets for classification tasks in a joint embedding space of image and text pairs. However, it showed negative transfer performance on standard datasets, e.g., BirdsNAP, RESISC45, and MNIST. In this paper, we propose ContextCLIP, a contextual and contrastive learning framework for the contextual alignment of image-text pairs by learning robust visual representations on Conceptual Captions dataset. Our framework was observed to improve the image-text alignment by aligning text and image representations contextually in the joint embedding space. ContextCLIP showed good qualitative performance for text-to-image retrieval tasks and enhanced classification accuracy. We evaluated our model quantitatively with zero-shot transfer and fine-tuning experiments on CIFAR-10, CIFAR-100, Birdsnap, RESISC45, and MNIST datasets for classification task.

cs.CV↗

arXiv2022

SCOTCH: An Efficient Secure Computation Framework for Secure Aggregation

Federated learning enables multiple data owners to jointly train a machine learning model without revealing their private datasets. However, a malicious aggregation server might use the model parameters to derive sensitive information about the training dataset used. To address such leakage, differential privacy and cryptographic techniques have been investigated in prior work, but these often result in large communication overheads or impact model performance. To mitigate this centralization of power, we propose SCOTCH, a decentralized m-party secure-computation framework for federated aggregation that deploys MPC primitives, such as secret sharing. Our protocol is simple, efficient, and provides strict privacy guarantees against curious aggregators or colluding data-owners with minimal communication overheads compared to other existing state-of-the-art privacy-preserving federated learning frameworks. We evaluate our framework by performing extensive experiments on multiple datasets with promising results. SCOTCH can train the standard MLP NN with the training dataset split amongst 3 participating users and 3 aggregating servers with 96.57% accuracy on MNIST, and 98.40% accuracy on the Extended MNIST (digits) dataset, while providing various optimizations.

cs.CR↗

arXiv2022

BEAS: Blockchain Enabled Asynchronous & Secure Federated Machine Learning

Federated Learning (FL) enables multiple parties to distributively train a ML model without revealing their private datasets. However, it assumes trust in the centralized aggregator which stores and aggregates model updates. This makes it prone to gradient tampering and privacy leakage by a malicious aggregator. Malicious parties can also introduce backdoors into the joint model by poisoning the training data or model gradients. To address these issues, we present BEAS, the first blockchain-based framework for N-party FL that provides strict privacy guarantees of training data using gradient pruning (showing improved differential privacy compared to existing noise and clipping based techniques). Anomaly detection protocols are used to minimize the risk of data-poisoning attacks, along with gradient pruning that is further used to limit the efficacy of model-poisoning attacks. We also define a novel protocol to prevent premature convergence in heterogeneous learning environments. We perform extensive experiments on multiple datasets with promising results: BEAS successfully prevents privacy leakage from dataset reconstruction attacks, and minimizes the efficacy of poisoning attacks. Moreover, it achieves an accuracy similar to centralized frameworks, and its communication and computation overheads scale linearly with the number of participants.

cs.CR↗

arXiv2021

Flatee: Federated Learning Across Trusted Execution Environments

Federated learning allows us to distributively train a machine learning model where multiple parties share local model parameters without sharing private data. However, parameter exchange may still leak information. Several approaches have been proposed to overcome this, based on multi-party computation, fully homomorphic encryption, etc.; many of these protocols are slow and impractical for real-world use as they involve a large number of cryptographic operations. In this paper, we propose the use of Trusted Execution Environments (TEE), which provide a platform for isolated execution of code and handling of data, for this purpose. We describe Flatee, an efficient privacy-preserving federated learning framework across TEEs, which considerably reduces training and communication time. Our framework can handle malicious parties (we do not natively solve adversarial data poisoning, though we describe a preliminary approach to handle this).

cs.CR↗

arXiv2021

S++: A Fast and Deployable Secure-Computation Framework for Privacy-Preserving Neural Network Training

We introduce S++, a simple, robust, and deployable framework for training a neural network (NN) using private data from multiple sources, using secret-shared secure function evaluation. In short, consider a virtual third party to whom every data-holder sends their inputs, and which computes the neural network: in our case, this virtual third party is actually a set of servers which individually learn nothing, even with a malicious (but non-colluding) adversary. Previous work in this area has been limited to just one specific activation function: ReLU, rendering the approach impractical for many use-cases. For the first time, we provide fast and verifiable protocols for all common activation functions and optimize them for running in a secret-shared manner. The ability to quickly, verifiably, and robustly compute exponentiation, softmax, sigmoid, etc., allows us to use previously written NNs without modification, vastly reducing developer effort and complexity of code. In recent times, ReLU has been found to converge much faster and be more computationally efficient as compared to non-linear functions like sigmoid or tanh. However, we argue that it would be remiss not to extend the mechanism to non-linear functions such as the logistic sigmoid, tanh, and softmax that are fundamental due to their ability to express outputs as probabilities and their universal approximation property. Their contribution in RNNs and a few recent advancements also makes them more relevant.

cs.CR↗

arXiv2021

Differential Euler: Designing a Neural Network approximator to solve the Chaotic Three Body Problem

The three body problem is a special case of the n body problem where one takes the initial positions and velocities of three point masses and attempts to predict their motion over time according to Newtonian laws of motion and universal gravitation. Though analytical solutions have been found for special cases, the general problem remains unsolved; the solutions that do exist are impractical. Fortunately, for many applications, we may not need to solve the problem completely, i.e., predicting with reasonable accuracy for some time steps, may be sufficient. Recently, Breen et al attempted to approximately solve the three body problem using a simple neural network. Although their methods appear to achieve some success in reducing the computational overhead, their model is extremely restricted, applying to a specialized 2D case. The authors do not provide explanations for critical decisions taken in their experimental design, no details on their model or architecture, and nor do they publish their code. Moreover, the model does not generalize well to unseen cases. In this paper, we propose a detailed experimental setup to determine the feasibility of using neural networks to solve the three body problem up to a certain number of time steps. We establish a benchmark on the dataset size and set an accuracy threshold to measure the viability of our results for practical applications. Then, we build our models according to the listed class of NNs using a dataset generated from standard numerical integrators. We gradually increase the complexity of our data set to determine whether NNs can learn a representation of the chaotic three body problem well enough to replace numerical integrators in real life scenarios.

cs.LG↗

arXiv2018

A Peculiarity in the Parity of Primes

We create a simple test for distinguishing between sets of primes and random numbers using just the sum-of-digits function. We find that the sum-of-the-digits of prime numbers does not have an equal probability of being odd or even. The authors know of no reason why prime numbers should bias themselves towards a particular parity in their sums of digits, but our empirical tests show a very strong bias; strong enough that we are able to devise a test to reliably differentiate between collections of prime numbers versus random numbers by looking only at their sums of digits. We are also able to create similar tests for products of primes. We are even able to test "tainted" sets with mixtures of primes and random numbers: as the percentage of (randomly chosen) prime numbers in a set of random numbers is varied, we get a reliable, linear change in our parity measure. For example, when we add up the digits of prime numbers in base 10, their sum is significantly more likely to be odd than even. This effect persists across base changes, although which parity is more common might change. Note that the last digit being odd in base 10 simply reverses the parity. We have tested this for the first fifty million primes -- not primes up to 50,000,000, but the first 50,000,000 prime numbers -- and have found that this effect persists, and does so in a predictable manner. The effect is quite significant; for 50,000,000 primes in base 10, the number of primes which have an odd sum-of-digits is about an order of magnitude farther away from the mean than expected. We have run multiple tests to try and understand the source of this bias, including investigating primes modulo random numbers and adjusting for Chebyshev's bias. None of these tests yielded any satisfactory explanation for this phenomenon.

math.GM↗

arXiv2015

Reuse It Or Lose It: More Efficient Secure Computation Through Reuse of Encrypted Values

Two-party secure function evaluation (SFE) has become significantly more feasible, even on resource-constrained devices, because of advances in server-aided computation systems. However, there are still bottlenecks, particularly in the input validation stage of a computation. Moreover, SFE research has not yet devoted sufficient attention to the important problem of retaining state after a computation has been performed so that expensive processing does not have to be repeated if a similar computation is done again. This paper presents PartialGC, an SFE system that allows the reuse of encrypted values generated during a garbled-circuit computation. We show that using PartialGC can reduce computation time by as much as 96% and bandwidth by as much as 98% in comparison with previous outsourcing schemes for secure computation. We demonstrate the feasibility of our approach with two sets of experiments, one in which the garbled circuit is evaluated on a mobile device and one in which it is evaluated on a server. We also use PartialGC to build a privacy-preserving "friend finder" application for Android. The reuse of previous inputs to allow stateful evaluation represents a new way of looking at SFE and further reduces computational barriers.

cs.CR↗