Masked Image Modeling (MIM) has garnered significant attention in self-supervised learning, thanks to its impressive capacity to learn scalable visual representations tailored for downstream tasks. However, images inherently contain abundant redundant information, leading the pixel-based MIM reconstruction process to focus excessively on finer details such as textures, thus prolonging training times unnecessarily. Addressing this challenge requires a shift towards a compact representation of features during MIM reconstruction. Frequency domain analysis provides a promising avenue for achieving compact image feature representation. In contrast to the commonly used Fourier transform, wavelet transform not only offers frequency information but also preserves spatial characteristics and multi-level features of the image. Additionally, the multi-level decomposition process of wavelet transformation aligns well with the hierarchical architecture of modern neural networks. In this study, we leverage wavelet transform as a tool for efficient representation learning to expedite the training process of MIM. Specifically, we conduct multi-level decomposition of images using wavelet transform, utilizing wavelet coefficients from different levels to construct distinct reconstruction targets representing various frequencies and scales. These reconstruction targets are then integrated into the MIM process, with adjustable weights assigned to prioritize the most crucial information. Extensive experiments demonstrate that our method achieves comparable or superior performance across various downstream tasks while exhibiting higher training efficiency.

The advancement of multi-modal large language models (MLLMs) has significantly enhanced their capability to process and understand diverse data types, integrating text, images, and other modalities. Despite their impressive performance, evaluating the robustness of these models remains challenging due to the difficulty of aligning their text-based responses with image classification labels. Traditional approaches rely on CLIP scores or other large language models as judges, but these methods lack scientific rigor and fail to capture robustness across different semantic levels. In this paper, we propose a novel evaluation metric that systematically assesses the robustness of MLLMs in image classification using WordNet’s hierarchical structure. Specifically, we parse the text descriptions generated by MLLMs to extract all nouns, then calculate their distances to the groundtruth label in WordNet as a similarity metric. The minimum distance among all nouns is used as the similarity score between the text description and the label. By using WordNet, we can also evaluate classification performance at different semantic levels. Through extensive experiments, we demonstrate that our WordNet-based evaluation metric offers a deeper understanding of MLLMs’ robustness, paving the way for more resilient and reliable models in real-world applications.

Deep Neural Networks (DNNs) often suffer from performance drops when training and test data distributions differ. Ensuring model generalization for Out-Of-Distribution (OOD) data is crucial, but current models still struggle with accuracy on such data. Recent studies have shown that regular or off-manifold adversarial examples as data augmentation improve OOD generalization. Building on this, we provide theoretical validation that on-manifold adversarial examples can enhance OOD generalization even more. However, generating these examples is challenging due to the complexity of real manifolds. To address this, we propose AdvWavAug, an on-manifold adversarial data augmentation method using a Wavelet module. This approach, based on the AdvProp training framework, leverages wavelet transformation to project an image into the wavelet domain and modifies it within the estimated data manifold. Experiments on various models and datasets, including ImageNet and its distorted versions, show that our method significantly improves model generalization, especially for OOD data.

Although pre-trained models such as Contrastive Language-Image Pre-Training (CLIP) show impressive generalization results, their robustness is still limited under Out-of-Distribution (OOD) scenarios. Instead of undesirably leveraging human annotation as commonly done, it is possible to leverage the visual understanding power of Multi-modal Large Language Models (MLLMs). However, MLLMs struggle with vision problems due to task incompatibility, thus hindering their effectiveness. In this paper, we propose to effectively leverage MLLMs via Machine Vision Therapy which aims to rectify erroneous predictions of specific vision models. By supervising vision models using MLLM predictions, visual robustness can be boosted in a nearly unsupervised manner. Moreover, we propose a Denoising In-Context Learning (DICL) strategy to solve the incompatibility issue. Concretely, by examining the noise probability of each example through a transition matrix, we construct an instruction containing a correct exemplar and a probable erroneous one, which enables MLLMs to detect and rectify the incorrect predictions of vision models. Under mild assumptions, we theoretically show that our DICL method is guaranteed to find the ground truth. Through extensive experiments on various OOD datasets, our method demonstrates powerful capabilities for enhancing visual robustness under many OOD scenarios.

3D articulated objects are inherently challenging for manipulation due to the varied geometries and intricate functionalities associated with articulated objects. Point-level affordance, which predicts the per-point actionable score and thus proposes the best point to interact with, has demonstrated excellent performance and generalization capabilities in articulated object manipulation. However, a significant challenge remains: while previous works use perfect point cloud generated in simulation, the models cannot directly apply to the noisy point cloud in the real-world. To tackle this challenge, we leverage the property of real-world scanned point cloud that, the point cloud becomes less noisy when the camera is closer to the object. Therefore, we propose a novel coarse-to-fine affordance learning pipeline to mitigate the effect of point cloud noise in two stages. In the first stage, we learn the affordance on the noisy far point cloud which includes the whole object to propose the approximated place to manipulate. Then, we move the camera in front of the approximated place, scan a less noisy point cloud containing precise local geometries for manipulation, and learn affordance on such point cloud to propose fine-grained final actions. The proposed method is thoroughly evaluated both using large-scale simulated noisy point clouds mimicking real-world scans, and in the real world scenarios, with superiority over existing methods, demonstrating the effectiveness in tackling the noisy real-world point cloud problem.

Despite the superior capabilities of Multimodal Large Language Models (MLLMs) across diverse tasks, they still face significant trustworthiness challenges. Yet, current literature on the assessment of trustworthy MLLMs remains limited, lacking a holistic evaluation to offer thorough insights into future improvements. In this work, we establish MultiTrust, the first comprehensive and unified benchmark on the trustworthiness of MLLMs across five primary aspects: truthfulness, safety, robustness, fairness, and privacy. Our benchmark employs a rigorous evaluation strategy that addresses both multimodal risks and cross-modal impacts, encompassing 32 diverse tasks with self-curated datasets. Extensive experiments with 21 modern MLLMs reveal some previously unexplored trustworthiness issues and risks, highlighting the complexities introduced by the multimodality and underscoring the necessity for advanced methodologies to enhance their reliability. For instance, typical proprietary models still struggle with the perception of visually confusing images and are vulnerable to multimodal jailbreaking and adversarial attacks; MLLMs are more inclined to disclose privacy in text and reveal ideological and cultural biases even when paired with irrelevant images in inference, indicating that the multimodality amplifies the internal risks from base LLMs. Additionally, we release a scalable toolbox for standardized trustworthiness research, aiming to facilitate future advancements in this important field. Code and resources are publicly available at: https://multi-trust.github.io/.

The recommendation performance of deep collaborative filtering models drops sharply under imperceptible adversarial perturbations. Some methods promote the robustness of recommendation systems by adversarial training. However, these methods only study shallow models and lack the exploration of deep models. Furthermore, the way these methods add adversarial noise to the weight parameters of users and items is not fully applicable to deep collaborative filtering models, because the adversarial noise is not sufficient to fully affect its network structure with multiple hidden layers. In this article, we propose a novel adversarial training framework, Random Layer-wise Adversarial Training (RAT), which trains a robust deep collaborative filtering model via adversarial noise propagation. Specifically, we inject adversarial noise into the output of the hidden layer in a random layer-wise manner. The adversarial noise propagates forward from the injected position to obtain more flexible model parameters during the adversarial training process. We validate the effectiveness of RAT on multilayer perceptron (MLP) and implement RAT on MLP-based and convolutional neural networks-based deep collaborative filtering models. Experiments on three publicly available datasets show that the deep collaborative filtering model trained by RAT not only defends against adversarial noise but also guarantees recommendation performance.

Face recognition is a prevailing authentication solution in numerous biometric applications. Physical adversarial attacks, as an important surrogate, can identify the weak-nesses of face recognition systems and evaluate their ro-bustness before deployed. However, most existing physical attacks are either detectable readily or ineffective against commercial recognition systems. The goal of this work is to develop a more reliable technique that can carry out an end-to-end evaluation of adversarial robustness for commercial systems. It requires that this technique can simultaneously deceive black-box recognition models and evade defensive mechanisms. To fulfill this, we design adversarial textured 3D meshes (AT3D) with an elaborate topology on a human face, which can be 3D-printed and pasted on the attacker's face to evade the defenses. However, the mesh-based op-timization regime calculates gradients in high-dimensional mesh space, and can be trapped into local optima with un-satisfactory transferability. To deviate from the mesh-based space, we propose to perturb the low-dimensional coefficient space based on 3D Morphable Model, which signifi-cantly improves black-box transferability meanwhile enjoying faster search efficiency and better visual quality. Exten-sive experiments in digital and physical scenarios show that our method effectively explores the security vulnerabilities of multiple popular commercial services, including three recognition A PIs, four anti-spoofing A PIs, two prevailing mobile phones and two automated access control systems.