DeepEraser: Deep Iterative Context Mining for Generic Text Eraser

In this work, we present DeepEraser, an effective deep network for generic text removal. DeepEraser utilizes a recurrent architecture that erases the text in an image via iterative operations. Our idea comes from the process of erasing pencil script, where the text area designated for removal is subject to continuous monitoring and the text is attenuated progressively, ensuring a thorough and clean erasure. Technically, at each iteration, an innovative erasing module is deployed, which not only explicitly aggregates the previous erasing progress but also mines additional semantic context to erase the target text. Through iterative refinements, the text regions are progressively replaced with more appropriate content and finally converge to a relatively accurate status. Furthermore, a custom mask generation strategy is introduced to improve the capability of DeepEraser for adaptive text removal, as opposed to indiscriminately removing all the text in an image. Our DeepEraser is notably compact with only 1.4M parameters and trained in an end-to-end manner. To verify its effectiveness, extensive experiments are conducted on several prevalent benchmarks, including SCUT-Syn, SCUT-EnsText, and Oxford Synthetic text dataset. The quantitative and qualitative results demonstrate the effectiveness of our DeepEraser over the state-of-the-art methods, as well as its strong generalization ability in custom mask text removal. The codes and pre-trained models are available at https://github.com/fh2019ustc/DeepEraser

BLO-SAM: Bi-level Optimization Based Overfitting-Preventing Finetuning of SAM

The Segment Anything Model (SAM), a foundation model pretrained on millions of images and segmentation masks, has significantly advanced semantic segmentation, a fundamental task in computer vision. Despite its strengths, SAM encounters two major challenges. Firstly, it struggles with segmenting specific objects autonomously, as it relies on users to manually input prompts like points or bounding boxes to identify targeted objects. Secondly, SAM faces challenges in excelling at specific downstream tasks, like medical imaging, due to a disparity between the distribution of its pretraining data, which predominantly consists of general-domain images, and the data used in downstream tasks. Current solutions to these problems, which involve finetuning SAM, often lead to overfitting, a notable issue in scenarios with very limited data, like in medical imaging. To overcome these limitations, we introduce BLO-SAM, which finetunes SAM based on bi-level optimization (BLO). Our approach allows for automatic image segmentation without the need for manual prompts, by optimizing a learnable prompt embedding. Furthermore, it significantly reduces the risk of overfitting by training the model's weight parameters and the prompt embedding on two separate subsets of the training dataset, each at a different level of optimization. We apply BLO-SAM to diverse semantic segmentation tasks in general and medical domains. The results demonstrate BLO-SAM's superior performance over various state-of-the-art image semantic segmentation methods.

Finer: Investigating and Enhancing Fine-Grained Visual Concept Recognition in Large Vision Language Models

Recent advances in instruction-tuned Large Vision-Language Models (LVLMs) have imbued the models with the ability to generate high-level, image-grounded explanations with ease. While such capability is largely attributed to the rich world knowledge contained within the Large Language Models (LLMs), our work reveals their shortcomings in fine-grained visual categorization (FGVC) across six different benchmark settings. Most recent state-of-the-art LVLMs like LLaVa-1.5, InstructBLIP and GPT-4V not only severely deteriorate in terms of classification performance, e.g., average drop of 65.58 in EM for Stanford Dogs for LLaVA-1.5, but also struggle to generate an accurate explanation with detailed attributes based on the concept that appears within an input image despite their capability to generate holistic image-level descriptions. In-depth analyses show that instruction-tuned LVLMs exhibit modality gap, showing discrepancy when given textual and visual inputs that correspond to the same concept, preventing the image modality from leveraging the rich parametric knowledge within the LLMs. In an effort to further the community's endeavor in this direction, we propose a multiple granularity attribute-centric evaluation benchmark, Finer, which aims to establish a ground to evaluate LVLMs' fine-grained visual comprehension ability and provide significantly improved explainability.

A Performance Evaluation of Filtered Delay Multiply and Sum Beamforming for Ultrasound Localization Microscopy: Preliminary Results

Ultrafast ultrasound localization microscopy (ULM), which has shown promising results in microvascular imaging, overcomes the typical trade-off between resolution and penetration depth. Combining ultrasound contrast agents and high frame rate imaging enables ULM to visualize microvasculature and quantify flow. However, the quality of the microvascular maps obtained depends on the signal-to-noise ratio of the received signals, image reconstruction techniques, and the microbubble (MB) localization and tracking algorithms used. Most reported research in ULM employs the conventional delay and sum (DAS) beamforming technique for image reconstruction despite its limited contrast and resolution. In this work, a filtered delay multiply and sum (F-DMAS) beamforming approach with non-steered plane wave transmit was employed for ULM, and its performance was compared with the conventional DAS-based approach for the different localization algorithms available in the Localization and Tracking Toolbox for Ultrasound Localization Microscopy. We also introduce two novel image quality measures that can overcome the limitations of conventional quality metrics that require suitable targets for evaluation. We also report the preliminary in-vitro investigation of F-DMAS with B-mode and power Doppler maps for microvascular imaging. The results are promising with enhanced contrast and lateral resolution, and suggest that further experimental studies are warranted.

SAR-AE-SFP: SAR Imagery Adversarial Example in Real Physics domain with Target Scattering Feature Parameters

Deep neural network-based Synthetic Aperture Radar (SAR) target recognition models are susceptible to adversarial examples. Current adversarial example generation methods for SAR imagery primarily operate in the 2D digital domain, known as image adversarial examples. Recent work, while considering SAR imaging scatter mechanisms, fails to account for the actual imaging process, rendering attacks in the three-dimensional physical domain infeasible, termed pseudo physics adversarial examples. To address these challenges, this paper proposes SAR-AE-SFP-Attack, a method to generate real physics adversarial examples by altering the scattering feature parameters of target objects. Specifically, we iteratively optimize the coherent energy accumulation of the target echo by perturbing the reflection coefficient and scattering coefficient in the scattering feature parameters of the three-dimensional target object, and obtain the adversarial example after echo signal processing and imaging processing in the RaySAR simulator. Experimental results show that compared to digital adversarial attack methods, SAR-AE-SFP Attack significantly improves attack efficiency on CNN-based models (over 30\%) and Transformer-based models (over 13\%), demonstrating significant transferability of attack effects across different models and perspectives.

SDGE: Stereo Guided Depth Estimation for 360$^\circ$ Camera Sets

Depth estimation is a critical technology in autonomous driving, and multi-camera systems are often used to achieve a 360$^\circ$ perception. These 360$^\circ$ camera sets often have limited or low-quality overlap regions, making multi-view stereo methods infeasible for the entire image. Alternatively, monocular methods may not produce consistent cross-view predictions. To address these issues, we propose the Stereo Guided Depth Estimation (SGDE) method, which enhances depth estimation of the full image by explicitly utilizing multi-view stereo results on the overlap. We suggest building virtual pinhole cameras to resolve the distortion problem of fisheye cameras and unify the processing for the two types of 360$^\circ$ cameras. For handling the varying noise on camera poses caused by unstable movement, the approach employs a self-calibration method to obtain highly accurate relative poses of the adjacent cameras with minor overlap. These enable the use of robust stereo methods to obtain high-quality depth prior in the overlap region. This prior serves not only as an additional input but also as pseudo-labels that enhance the accuracy of depth estimation methods and improve cross-view prediction consistency. The effectiveness of SGDE is evaluated on one fisheye camera dataset, Synthetic Urban, and two pinhole camera datasets, DDAD and nuScenes. Our experiments demonstrate that SGDE is effective for both supervised and self-supervised depth estimation, and highlight the potential of our method for advancing downstream autonomous driving technologies, such as 3D object detection and occupancy prediction.

Less is More: Fewer Interpretable Region via Submodular Subset Selection

Image attribution algorithms aim to identify important regions that are highly relevant to model decisions. Although existing attribution solutions can effectively assign importance to target elements, they still face the following challenges: 1) existing attribution methods generate inaccurate small regions thus misleading the direction of correct attribution, and 2) the model cannot produce good attribution results for samples with wrong predictions. To address the above challenges, this paper re-models the above image attribution problem as a submodular subset selection problem, aiming to enhance model interpretability using fewer regions. To address the lack of attention to local regions, we construct a novel submodular function to discover more accurate small interpretation regions. To enhance the attribution effect for all samples, we also impose four different constraints on the selection of sub-regions, i.e., confidence, effectiveness, consistency, and collaboration scores, to assess the importance of various subsets. Moreover, our theoretical analysis substantiates that the proposed function is in fact submodular. Extensive experiments show that the proposed method outperforms SOTA methods on two face datasets (Celeb-A and VGG-Face2) and one fine-grained dataset (CUB-200-2011). For correctly predicted samples, the proposed method improves the Deletion and Insertion scores with an average of 4.9% and 2.5% gain relative to HSIC-Attribution. For incorrectly predicted samples, our method achieves gains of 81.0% and 18.4% compared to the HSIC-Attribution algorithm in the average highest confidence and Insertion score respectively. The code is released at https://github.com/RuoyuChen10/SMDL-Attribution.

Real-Time Estimation of Relative Pose for UAVs Using a Dual-Channel Feature Association

Leveraging multiple cameras on Unmanned Aerial Vehicles (UAVs) to form a variable-baseline stereo camera for collaborative perception is highly promising. The critical steps include high-rate cross-camera feature association and frame-rate relative pose estimation of multiple UAVs. To accelerate the feature association rate to match the frame rate, we propose a dual-channel structure to decouple the time-consuming feature detection and match from the high-rate image stream. The novel design of periodic guidance and fast prediction effectively utilizes each image frame to achieve a frame-rate feature association. Real-world experiments are executed using SuperPoint and SuperGlue on the NVIDIA NX 8G platform with a 30 Hz image stream. Using single-channel SuperPoint and SuperGlue can only achieve 13 Hz feature association. The proposed dual-channel method can improve the rate of feature association from 13 Hz to 30 Hz, supporting the frame-rate requirement. To accommodate the proposed feature association, we develop a Multi-State Constrained Kalman Filter (MSCKF)-based relative pose estimator in the back-end by fusing the local odometry from two UAVs together with the measurements of common features. Experiments show that the dual-channel feature association improves the rate of visual observation and enhances the real-time performance of back-end estimator compared to the existing methods. Video - https://youtu.be/UBAR1iP0GPk Supplementary video - https://youtu.be/nPq8EpVzJZM

Generative AI in Vision: A Survey on Models, Metrics and Applications

Generative AI models have revolutionized various fields by enabling the creation of realistic and diverse data samples. Among these models, diffusion models have emerged as a powerful approach for generating high-quality images, text, and audio. This survey paper provides a comprehensive overview of generative AI diffusion and legacy models, focusing on their underlying techniques, applications across different domains, and their challenges. We delve into the theoretical foundations of diffusion models, including concepts such as denoising diffusion probabilistic models (DDPM) and score-based generative modeling. Furthermore, we explore the diverse applications of these models in text-to-image, image inpainting, and image super-resolution, along with others, showcasing their potential in creative tasks and data augmentation. By synthesizing existing research and highlighting critical advancements in this field, this survey aims to provide researchers and practitioners with a comprehensive understanding of generative AI diffusion and legacy models and inspire future innovations in this exciting area of artificial intelligence.

Approximations to the Fisher Information Metric of Deep Generative Models for Out-Of-Distribution Detection

Likelihood-based deep generative models such as score-based diffusion models and variational autoencoders are state-of-the-art machine learning models approximating high-dimensional distributions of data such as images, text, or audio. One of many downstream tasks they can be naturally applied to is out-of-distribution (OOD) detection. However, seminal work by Nalisnick et al. which we reproduce showed that deep generative models consistently infer higher log-likelihoods for OOD data than data they were trained on, marking an open problem. In this work, we analyse using the gradient of a data point with respect to the parameters of the deep generative model for OOD detection, based on the simple intuition that OOD data should have larger gradient norms than training data. We formalise measuring the size of the gradient as approximating the Fisher information metric. We show that the Fisher information matrix (FIM) has large absolute diagonal values, motivating the use of chi-square distributed, layer-wise gradient norms as features. We combine these features to make a simple, model-agnostic and hyperparameter-free method for OOD detection which estimates the joint density of the layer-wise gradient norms for a given data point. We find that these layer-wise gradient norms are weakly correlated, rendering their combined usage informative, and prove that the layer-wise gradient norms satisfy the principle of (data representation) invariance. Our empirical results indicate that this method outperforms the Typicality test for most deep generative models and image dataset pairings.