The Vision Transformer architecture is a deep learning model inspired by the success of the Transformer model in Natural Language Processing. However, the self-attention mechanism, large number of parameters, and the requirement for a substantial amount of training data still make Vision Transformers computationally burdensome. In this research, we investigate the possibility of applying Sparse Regularization to Vision Transformers and the impact of Pruning, either after Sparse Regularization or without it, on the trade-off between performance and efficiency. To accomplish this, we apply Sparse Regularization and Pruning methods to the Vision Transformer architecture for image classification tasks on the CIFAR-10, CIFAR-100, and ImageNet-100 datasets. The training process for the Vision Transformer model consists of two parts: pre-training and fine-tuning. Pre-training utilizes ImageNet21K data, followed by fine-tuning for 20 epochs. The results show that when testing with CIFAR-100 and ImageNet-100 data, models with Sparse Regularization can increase accuracy by 0.12%. Furthermore, applying pruning to models with Sparse Regularization yields even better results. Specifically, it increases the average accuracy by 0.568% on CIFAR-10 data, 1.764% on CIFAR-100, and 0.256% on ImageNet-100 data compared to pruning models without Sparse Regularization. Code can be accesed here: https://github.com/yogiprsty/Sparse-ViT
In response to growing FinTech competition and the need for improved operational efficiency, this research focuses on understanding the potential of advanced document analytics, particularly using multimodal models, in banking processes. We perform a comprehensive analysis of the diverse banking document landscape, highlighting the opportunities for efficiency gains through automation and advanced analytics techniques in the customer business. Building on the rapidly evolving field of natural language processing (NLP), we illustrate the potential of models such as LayoutXLM, a cross-lingual, multimodal, pre-trained model, for analyzing diverse documents in the banking sector. This model performs a text token classification on German company register extracts with an overall F1 score performance of around 80\%. Our empirical evidence confirms the critical role of layout information in improving model performance and further underscores the benefits of integrating image information. Interestingly, our study shows that over 75% F1 score can be achieved with only 30% of the training data, demonstrating the efficiency of LayoutXLM. Through addressing state-of-the-art document analysis frameworks, our study aims to enhance process efficiency and demonstrate the real-world applicability and benefits of multimodal models within banking.
Convolutional Neural Networks (CNN) are known to exhibit poor generalization performance under distribution shifts. Their generalization have been studied extensively, and one line of work approaches the problem from a frequency-centric perspective. These studies highlight the fact that humans and CNNs might focus on different frequency components of an image. First, inspired by these observations, we propose a simple yet effective data augmentation method HybridAugment that reduces the reliance of CNNs on high-frequency components, and thus improves their robustness while keeping their clean accuracy high. Second, we propose HybridAugment++, which is a hierarchical augmentation method that attempts to unify various frequency-spectrum augmentations. HybridAugment++ builds on HybridAugment, and also reduces the reliance of CNNs on the amplitude component of images, and promotes phase information instead. This unification results in competitive to or better than state-of-the-art results on clean accuracy (CIFAR-10/100 and ImageNet), corruption benchmarks (ImageNet-C, CIFAR-10-C and CIFAR-100-C), adversarial robustness on CIFAR-10 and out-of-distribution detection on various datasets. HybridAugment and HybridAugment++ are implemented in a few lines of code, does not require extra data, ensemble models or additional networks.
Visual question answering requires a system to provide an accurate natural language answer given an image and a natural language question. However, it is widely recognized that previous generic VQA methods often exhibit a tendency to memorize biases present in the training data rather than learning proper behaviors, such as grounding images before predicting answers. Therefore, these methods usually achieve high in-distribution but poor out-of-distribution performance. In recent years, various datasets and debiasing methods have been proposed to evaluate and enhance the VQA robustness, respectively. This paper provides the first comprehensive survey focused on this emerging fashion. Specifically, we first provide an overview of the development process of datasets from in-distribution and out-of-distribution perspectives. Then, we examine the evaluation metrics employed by these datasets. Thirdly, we propose a typology that presents the development process, similarities and differences, robustness comparison, and technical features of existing debiasing methods. Furthermore, we analyze and discuss the robustness of representative vision-and-language pre-training models on VQA. Finally, through a thorough review of the available literature and experimental analysis, we discuss the key areas for future research from various viewpoints.
This work presents an approach for automating the discretization and approximation procedures in constructing digital representations of composites from Micro-CT images featuring intricate microstructures. The proposed method is guided by the Support Vector Machine (SVM) classification, offering an effective approach for discretizing microstructural images. An SVM soft margin training process is introduced as a classification of heterogeneous material points, and image segmentation is accomplished by identifying support vectors through a local regularized optimization problem. In addition, an Interface-Modified Reproducing Kernel Particle Method (IM-RKPM) is proposed for appropriate approximations of weak discontinuities across material interfaces. The proposed method modifies the smooth kernel functions with a regularized heavy-side function concerning the material interfaces to alleviate Gibb's oscillations. This IM-RKPM is formulated without introducing duplicated degrees of freedom associated with the interface nodes commonly needed in the conventional treatments of weak discontinuities in the meshfree methods. Moreover, IM-RKPM can be implemented with various domain integration techniques, such as Stabilized Conforming Nodal Integration (SCNI). The extension of the proposed method to 3-dimension is straightforward, and the effectiveness of the proposed method is validated through the image-based modeling of polymer-ceramic composite microstructures.
Large language models are few-shot learners that can solve diverse tasks from a handful of demonstrations. This implicit understanding of tasks suggests that the attention mechanisms over word tokens may play a role in analogical reasoning. In this work, we investigate whether analogical reasoning can enable in-context composition over composable elements of visual stimuli. First, we introduce a suite of three benchmarks to test the generalization properties of a visual in-context learner. We formalize the notion of an analogy-based in-context learner and use it to design a meta-learning framework called Im-Promptu. Whereas the requisite token granularity for language is well established, the appropriate compositional granularity for enabling in-context generalization in visual stimuli is usually unspecified. To this end, we use Im-Promptu to train multiple agents with different levels of compositionality, including vector representations, patch representations, and object slots. Our experiments reveal tradeoffs between extrapolation abilities and the degree of compositionality, with non-compositional representations extending learned composition rules to unseen domains but performing poorly on combinatorial tasks. Patch-based representations require patches to contain entire objects for robust extrapolation. At the same time, object-centric tokenizers coupled with a cross-attention module generate consistent and high-fidelity solutions, with these inductive biases being particularly crucial for compositional generalization. Lastly, we demonstrate a use case of Im-Promptu as an intuitive programming interface for image generation.
Removing multiple degradations, such as haze, rain, and blur, from real-world images poses a challenging and illposed problem. Recently, unified models that can handle different degradations have been proposed and yield promising results. However, these approaches focus on synthetic images and experience a significant performance drop when applied to realworld images. In this paper, we introduce Uni-Removal, a twostage semi-supervised framework for addressing the removal of multiple degradations in real-world images using a unified model and parameters. In the knowledge transfer stage, Uni-Removal leverages a supervised multi-teacher and student architecture in the knowledge transfer stage to facilitate learning from pretrained teacher networks specialized in different degradation types. A multi-grained contrastive loss is introduced to enhance learning from feature and image spaces. In the domain adaptation stage, unsupervised fine-tuning is performed by incorporating an adversarial discriminator on real-world images. The integration of an extended multi-grained contrastive loss and generative adversarial loss enables the adaptation of the student network from synthetic to real-world domains. Extensive experiments on real-world degraded datasets demonstrate the effectiveness of our proposed method. We compare our Uni-Removal framework with state-of-the-art supervised and unsupervised methods, showcasing its promising results in real-world image dehazing, deraining, and deblurring simultaneously.
Comparing different age estimation methods poses a challenge due to the unreliability of published results, stemming from inconsistencies in the benchmarking process. Previous studies have reported continuous performance improvements over the past decade using specialized methods; however, our findings challenge these claims. We argue that, for age estimation tasks outside of the low-data regime, designing specialized methods is unnecessary, and the standard approach of utilizing cross-entropy loss is sufficient. This paper aims to address the benchmark shortcomings by evaluating state-of-the-art age estimation methods in a unified and comparable setting. We systematically analyze the impact of various factors, including facial alignment, facial coverage, image resolution, image representation, model architecture, and the amount of data on age estimation results. Surprisingly, these factors often exert a more significant influence than the choice of the age estimation method itself. We assess the generalization capability of each method by evaluating the cross-dataset performance for publicly available age estimation datasets. The results emphasize the importance of using consistent data preprocessing practices and establishing standardized benchmarks to ensure reliable and meaningful comparisons. The source code is available at https://github.com/paplhjak/Facial-Age-Estimation-Benchmark.
Adaptive gradient-based optimizers, particularly Adam, have left their mark in training large-scale deep learning models. The strength of such optimizers is that they exhibit fast convergence while being more robust to hyperparameter choice. However, they often generalize worse than non-adaptive methods. Recent studies have tied this performance gap to flat minima selection: adaptive methods tend to find solutions in sharper basins of the loss landscape, which in turn hurts generalization. To overcome this issue, we propose a new memory-augmented version of Adam that promotes exploration towards flatter minima by using a buffer of critical momentum terms during training. Intuitively, the use of the buffer makes the optimizer overshoot outside the basin of attraction if it is not wide enough. We empirically show that our method improves the performance of several variants of Adam on standard supervised language modelling and image classification tasks.
Interactive Image Retrieval (IIR) aims to retrieve images that are generally similar to the reference image but under the requested text modification. The existing methods usually concatenate or sum the features of image and text simply and roughly, which, however, is difficult to precisely change the local semantics of the image that the text intends to modify. To solve this problem, we propose a Language Guided Local Infiltration (LGLI) system, which fully utilizes the text information and penetrates text features into image features as much as possible. Specifically, we first propose a Language Prompt Visual Localization (LPVL) module to generate a localization mask which explicitly locates the region (semantics) intended to be modified. Then we introduce a Text Infiltration with Local Awareness (TILA) module, which is deployed in the network to precisely modify the reference image and generate image-text infiltrated representation. Extensive experiments on various benchmark databases validate that our method outperforms most state-of-the-art IIR approaches.