arXiv subjects

Gaurav Mittal

Publications and source records attributed to Gaurav Mittal.

At least 37 records · Page 2Linked to original sources

arXiv2023

Rethinking Multimodal Content Moderation from an Asymmetric Angle with Mixed-modality

There is a rapidly growing need for multimodal content moderation (CM) as more and more content on social media is multimodal in nature. Existing unimodal CM systems may fail to catch harmful content that crosses modalities (e.g., memes or videos), which may lead to severe consequences. In this paper, we present a novel CM model, Asymmetric Mixed-Modal Moderation (AM3), to target multimodal and unimodal CM tasks. Specifically, to address the asymmetry in semantics between vision and language, AM3 has a novel asymmetric fusion architecture that is designed to not only fuse the common knowledge in both modalities but also to exploit the unique information in each modality. Unlike previous works that focus on representing the two modalities into a similar feature space while overlooking the intrinsic difference between the information conveyed in multimodality and in unimodality (asymmetry in modalities), we propose a novel cross-modality contrastive loss to learn the unique knowledge that only appears in multimodality. This is critical as some harmful intent may only be conveyed through the intersection of both modalities. With extensive experiments, we show that AM3 outperforms all existing state-of-the-art methods on both multimodal and unimodal CM benchmarks.

cs.CV↗

arXiv2023

Rule By Example: Harnessing Logical Rules for Explainable Hate Speech Detection

Classic approaches to content moderation typically apply a rule-based heuristic approach to flag content. While rules are easily customizable and intuitive for humans to interpret, they are inherently fragile and lack the flexibility or robustness needed to moderate the vast amount of undesirable content found online today. Recent advances in deep learning have demonstrated the promise of using highly effective deep neural models to overcome these challenges. However, despite the improved performance, these data-driven models lack transparency and explainability, often leading to mistrust from everyday users and a lack of adoption by many platforms. In this paper, we present Rule By Example (RBE): a novel exemplar-based contrastive learning approach for learning from logical rules for the task of textual content moderation. RBE is capable of providing rule-grounded predictions, allowing for more explainable and customizable predictions compared to typical deep learning-based approaches. We demonstrate that our approach is capable of learning rich rule embedding representations using only a few data examples. Experimental results on 3 popular hate speech classification datasets show that RBE is able to outperform state-of-the-art deep learning classifiers as well as the use of rules in both supervised and unsupervised settings while providing explainable model predictions via rule-grounding.

cs.CL↗

arXiv2022

Convergence rates of nonlinear inverse problems in Banach spaces: conditional stability and weaker norms

In this short note, we formulate the convergence rates of the well known Tikhonov regularization scheme for solving the nonlinear ill-posed problems in Banach spaces. For deriving the convergence rates, we employ the novel smoothness concept of conditional stability estimates in terms of weaker norms. Moreover, we show that our results are applicable on two ill-posed inverse problems.

math.NA↗

arXiv2022

BATMAN: Bilateral Attention Transformer in Motion-Appearance Neighboring Space for Video Object Segmentation

Video Object Segmentation (VOS) is fundamental to video understanding. Transformer-based methods show significant performance improvement on semi-supervised VOS. However, existing work faces challenges segmenting visually similar objects in close proximity of each other. In this paper, we propose a novel Bilateral Attention Transformer in Motion-Appearance Neighboring space (BATMAN) for semi-supervised VOS. It captures object motion in the video via a novel optical flow calibration module that fuses the segmentation mask with optical flow estimation to improve within-object optical flow smoothness and reduce noise at object boundaries. This calibrated optical flow is then employed in our novel bilateral attention, which computes the correspondence between the query and reference frames in the neighboring bilateral space considering both motion and appearance. Extensive experiments validate the effectiveness of BATMAN architecture by outperforming all existing state-of-the-art on all four popular VOS benchmarks: Youtube-VOS 2019 (85.0%), Youtube-VOS 2018 (85.3%), DAVIS 2017Val/Testdev (86.2%/82.2%), and DAVIS 2016 (92.5%).

cs.CV↗

arXiv2022

GateHUB: Gated History Unit with Background Suppression for Online Action Detection

Online action detection is the task of predicting the action as soon as it happens in a streaming video. A major challenge is that the model does not have access to the future and has to solely rely on the history, i.e., the frames observed so far, to make predictions. It is therefore important to accentuate parts of the history that are more informative to the prediction of the current frame. We present GateHUB, Gated History Unit with Background Suppression, that comprises a novel position-guided gated cross-attention mechanism to enhance or suppress parts of the history as per how informative they are for current frame prediction. GateHUB further proposes Future-augmented History (FaH) to make history features more informative by using subsequently observed frames when available. In a single unified framework, GateHUB integrates the transformer's ability of long-range temporal modeling and the recurrent model's capacity to selectively encode relevant information. GateHUB also introduces a background suppression objective to further mitigate false positive background frames that closely resemble the action frames. Extensive validation on three benchmark datasets, THUMOS, TVSeries, and HDD, demonstrates that GateHUB significantly outperforms all existing methods and is also more efficient than the existing best work. Furthermore, a flow-free version of GateHUB is able to achieve higher or close accuracy at 2.8x higher frame rate compared to all existing methods that require both RGB and optical flow information for prediction.

cs.CV↗

arXiv2021

MUSE: Feature Self-Distillation with Mutual Information and Self-Information

We present a novel information-theoretic approach to introduce dependency among features of a deep convolutional neural network (CNN). The core idea of our proposed method, called MUSE, is to combine MUtual information and SElf-information to jointly improve the expressivity of all features extracted from different layers in a CNN. We present two variants of the realization of MUSE -- Additive Information and Multiplicative Information. Importantly, we argue and empirically demonstrate that MUSE, compared to other feature discrepancy functions, is a more functional proxy to introduce dependency and effectively improve the expressivity of all features in the knowledge distillation framework. MUSE achieves superior performance over a variety of popular architectures and feature discrepancy functions for self-distillation and online distillation, and performs competitively with the state-of-the-art methods for offline distillation. MUSE is also demonstrably versatile that enables it to be easily extended to CNN-based models on tasks other than image classification such as object detection.

cs.CV↗

arXiv2021

Unsupervised Few-Shot Action Recognition via Action-Appearance Aligned Meta-Adaptation

We present MetaUVFS as the first Unsupervised Meta-learning algorithm for Video Few-Shot action recognition. MetaUVFS leverages over 550K unlabeled videos to train a two-stream 2D and 3D CNN architecture via contrastive learning to capture the appearance-specific spatial and action-specific spatio-temporal video features respectively. MetaUVFS comprises a novel Action-Appearance Aligned Meta-adaptation (A3M) module that learns to focus on the action-oriented video features in relation to the appearance features via explicit few-shot episodic meta-learning over unsupervised hard-mined episodes. Our action-appearance alignment and explicit few-shot learner conditions the unsupervised training to mimic the downstream few-shot task, enabling MetaUVFS to significantly outperform all unsupervised methods on few-shot benchmarks. Moreover, unlike previous few-shot action recognition methods that are supervised, MetaUVFS needs neither base-class labels nor a supervised pretrained backbone. Thus, we need to train MetaUVFS just once to perform competitively or sometimes even outperform state-of-the-art supervised methods on popular HMDB51, UCF101, and Kinetics100 few-shot datasets.

cs.CV↗

arXiv2021

On Adversarial Robustness: A Neural Architecture Search perspective

Adversarial robustness of deep learning models has gained much traction in the last few years. Various attacks and defenses are proposed to improve the adversarial robustness of modern-day deep learning architectures. While all these approaches help improve the robustness, one promising direction for improving adversarial robustness is unexplored, i.e., the complex topology of the neural network architecture. In this work, we address the following question: Can the complex topology of a neural network give adversarial robustness without any form of adversarial training?. We answer this empirically by experimenting with different hand-crafted and NAS-based architectures. Our findings show that, for small-scale attacks, NAS-based architectures are more robust for small-scale datasets and simple tasks than hand-crafted architectures. However, as the size of the dataset or the complexity of task increases, hand-crafted architectures are more robust than NAS-based architectures. Our work is the first large-scale study to understand adversarial robustness purely from an architectural perspective. Our study shows that random sampling in the search space of DARTS (a popular NAS method) with simple ensembling can improve the robustness to PGD attack by nearly~12\%. We show that NAS, which is popular for achieving SoTA accuracy, can provide adversarial accuracy as a free add-on without any form of adversarial training. Our results show that leveraging the search space of NAS methods with methods like ensembles can be an excellent way to achieve adversarial robustness without any form of adversarial training. We also introduce a metric that can be used to calculate the trade-off between clean accuracy and adversarial robustness. Code and pre-trained models will be made available at \url{https://github.com/tdchaitanya/nas-robustness}

cs.LG↗

arXiv2021

A novel two-point gradient method for Regularization of inverse problems in Banach spaces

In this paper, we introduce a novel two-point gradient method for solving the ill-posed problems in Banach spaces and study its convergence analysis. The method is based on the well known iteratively regularized Landweber iteration method together with an extrapolation strategy. The general formulation of iteratively regularized Landweber iteration method in Banach spaces excludes the use of certain functions such as total variation like penalty functionals, $L^1$ functions etc. The novel scheme presented in this paper allows to use such non-smooth penalty terms that can be helpful in practical applications involving the reconstruction of several important features of solutions such as piecewise constancy and sparsity. We carefully discuss the choices for important parameters, such as combination parameters and step sizes involved in the design of the method. Additionally, we discuss an example to validate our assumptions.

math.NA↗

arXiv2021

A Vector Space Approach to Generate Dynamic Keys for Hill Cipher

In this paper, a variant of the Hill cipher is proposed. In the classical Hill cipher, an invertible matrix is used for encryption but the scheme is vulnerable to the known-plaintext attack which can reveal the matrix. In our proposed cryptosystem, each plaintext block is encrypted by a new invertible key matrix that thwarts the known-plaintext attack. To generate the invertible matrices which serve as the dynamic keys we make use of the vector spaces, randomly generated basis and non-singular linear transformation. Resulting cipher is secure against the known-plaintext attack.

cs.CR↗

arXiv2020

Non trivial idempotents of the matrix rings over polynomial ring Z_{pqr}[x]

In this paper, we study the non trivial idempotents of the $2 \times 2$ matrix ring over the polynomial ring $\mathbb{Z}_{pqr}[x]$ for distinct primes $p, q $ and $r$ greater than $3$. We have classified all the idempotents of this matrix ring into several classes such that any idempotent must belong to one of these classes. This work is extension of the work done in $[1]$.

math.RA↗

arXiv2020

Iteratively regularized landweber iteration method: Convergence analysis via Hölder Stability

In this paper, the local convergence of Iteratively regularized Landweber iteration method is investigated for solving non-linear inverse problems in Banach spaces. Our analysis mainly relies on the assumption that the inverse mapping satisfies the Hölder stability estimate locally. We consider both noisy as well as non-noisy data in our analysis. Under the a-priori choice of stopping index for noisy data, we show that the iterates remain in a certain ball around exact solution and obtain the convergence rates. The convergence of the Iteratively regularized Landweber iterates to the exact solution is shown under certain assumptions in the case of non-noisy data and as a by-product, under different conditions, two different convergence rates are obtained.

math.NA↗

arXiv2020

Convergence rates for inverse Problems in Hilbert spaces: A Comparative Study

In this paper, we apply a new kind of smoothness concept, i.e. Hölder stability estimates for the determination of convergence rates of Tikhonov regularization for linear and non-linear inverse problems in Hilbert spaces. For linear inverse problems, we obtain the convergence rates without incorporating the classical concept of spectral theory and for non-linear inverse problems, we obtain the convergence rates without incorporating any additional non-linearity estimate. Further, we employ the smoothness concept of inhomogeneous variational inequalities to deduce the convergence rates for non-linear inverse problems. In addition to Tikhonov regularization, we also consider Lavrentiev's regularization method for non-linear inverse problems and determine its convergence rates by incorporating the Hölder stability estimates as well as inhomogeneous variational inequalities. And finally, we discuss the co-action between the variational inequalities and the Hölder stability estimates.

math.NA↗

arXiv2020

Group ring based public key cryptosystems

In this paper, we propose two cryptosystems based on group rings and existing cryptosystem. First one is Elliptic ElGamal type group ring public key cryptosystem whose security is greater than security of cryptosystems based on elliptic curves discrete logarithmic problem (ECDLP). Second is ElGamal type group ring public key cryptosystem, which is analogous to ElGamal public key cryptosystem but has comparatively greater security. Examples are also given for both the proposed cryptosystems.

math.GR↗

arXiv2020

HyperSTAR: Task-Aware Hyperparameters for Deep Networks

While deep neural networks excel in solving visual recognition tasks, they require significant effort to find hyperparameters that make them work optimally. Hyperparameter Optimization (HPO) approaches have automated the process of finding good hyperparameters but they do not adapt to a given task (task-agnostic), making them computationally inefficient. To reduce HPO time, we present HyperSTAR (System for Task Aware Hyperparameter Recommendation), a task-aware method to warm-start HPO for deep neural networks. HyperSTAR ranks and recommends hyperparameters by predicting their performance conditioned on a joint dataset-hyperparameter space. It learns a dataset (task) representation along with the performance predictor directly from raw images in an end-to-end fashion. The recommendations, when integrated with an existing HPO method, make it task-aware and significantly reduce the time to achieve optimal performance. We conduct extensive experiments on 10 publicly available large-scale image classification datasets over two different network architectures, validating that HyperSTAR evaluates 50% less configurations to achieve the best performance compared to existing methods. We further demonstrate that HyperSTAR makes Hyperband (HB) task-aware, achieving the optimal accuracy in just 25% of the budget required by both vanilla HB and Bayesian Optimized HB~(BOHB).

cs.CV↗

arXiv2019

A note on an inverse problem in algebra

In this paper, we discuss the inverse problem of determining a semisimple group algebra from the knowledge of rings of the type sum_{t=1}^s M_{n_t}(Ft), where j is an arbitrary integer and F_t is finite field for each t, and show that it is ill-posed. After then, we define the concept of completeness of the rings of the type sum_{t=1}^s M_{n_t}(Ft) to pose a well-posed inverse problem and propose a conjecture in this direction.

math.RA↗

arXiv2019

Animating Face using Disentangled Audio Representations

All previous methods for audio-driven talking head generation assume the input audio to be clean with a neutral tone. As we show empirically, one can easily break these systems by simply adding certain background noise to the utterance or changing its emotional tone (to such as sad). To make talking head generation robust to such variations, we propose an explicit audio representation learning framework that disentangles audio sequences into various factors such as phonetic content, emotional tone, background noise and others. We conduct experiments to validate that conditioned on disentangled content representation, the generated mouth movement by our model is significantly more accurate than previous approaches (without disentangled learning) in the presence of noise and emotional variations. We further demonstrate that our framework is compatible with current state-of-the-art approaches by replacing their original audio learning component with ours. To our best knowledge, this is the first work which improves the performance of talking head generation from disentangled audio representation perspective, which is important for many real-world applications.

cs.CV↗

arXiv2019

Interactive Image Generation Using Scene Graphs

Recent years have witnessed some exciting developments in the domain of generating images from scene-based text descriptions. These approaches have primarily focused on generating images from a static text description and are limited to generating images in a single pass. They are unable to generate an image interactively based on an incrementally additive text description (something that is more intuitive and similar to the way we describe an image). We propose a method to generate an image incrementally based on a sequence of graphs of scene descriptions (scene-graphs). We propose a recurrent network architecture that preserves the image content generated in previous steps and modifies the cumulative image as per the newly provided scene information. Our model utilizes Graph Convolutional Networks (GCN) to cater to variable-sized scene graphs along with Generative Adversarial image translation networks to generate realistic multi-object images without needing any intermediate supervision during training. We experiment with Coco-Stuff dataset which has multi-object images along with annotations describing the visual scene and show that our model significantly outperforms other approaches on the same dataset in generating visually consistent images for incrementally growing scene graphs.

cs.CV↗