Recent advancements in biomedical image analysis have been significantly driven by the Segment Anything Model (SAM). This transformative technology, originally developed for general-purpose computer vision, has found rapid application in medical image processing. Within the last year, marked by over 100 publications, SAM has demonstrated its prowess in zero-shot learning adaptations for medical imaging. The fundamental premise of SAM lies in its capability to segment or identify objects in images without prior knowledge of the object type or imaging modality. This approach aligns well with tasks achievable by the human visual system, though its application in non-biological vision contexts remains more theoretically challenging. A notable feature of SAM is its ability to adjust segmentation according to a specified resolution scale or area of interest, akin to semantic priming. This adaptability has spurred a wave of creativity and innovation in applying SAM to medical imaging. Our review focuses on the period from April 1, 2023, to September 30, 2023, a critical first six months post-initial publication. We examine the adaptations and integrations of SAM necessary to address longstanding clinical challenges, particularly in the context of 33 open datasets covered in our analysis. While SAM approaches or achieves state-of-the-art performance in numerous applications, it falls short in certain areas, such as segmentation of the carotid artery, adrenal glands, optic nerve, and mandible bone. Our survey delves into the innovative techniques where SAM's foundational approach excels and explores the core concepts in translating and applying these models effectively in diverse medical imaging scenarios.
In this study, we consider the reliability assessment of anomaly detection (AD) using Variational Autoencoder (VAE). Over the last decade, VAE-based AD has been actively studied in various perspective, from method development to applied research. However, when the results of ADs are used in high-stakes decision-making, such as in medical diagnosis, it is necessary to ensure the reliability of the detected anomalies. In this study, we propose the VAE-AD Test as a method for quantifying the statistical reliability of VAE-based AD within the framework of statistical testing. Using the VAE-AD Test, the reliability of the anomaly regions detected by a VAE can be quantified in the form of p-values. This means that if an anomaly is declared when the p-value is below a certain threshold, it is possible to control the probability of false detection to a desired level. Since the VAE-AD Test is constructed based on a new statistical inference framework called selective inference, its validity is theoretically guaranteed in finite samples. To demonstrate the validity and effectiveness of the proposed VAE-AD Test, numerical experiments on artificial data and applications to brain image analysis are conducted.
This paper introduces a novel nonparametric framework for data imputation, coined multilinear kernel regression and imputation via the manifold assumption (MultiL-KRIM). Motivated by manifold learning, MultiL-KRIM models data features as a point cloud located in or close to a user-unknown smooth manifold embedded in a reproducing kernel Hilbert space. Unlike typical manifold-learning routes, which seek low-dimensional patterns via regularizers based on graph-Laplacian matrices, MultiL-KRIM builds instead on the intuitive concept of tangent spaces to manifolds and incorporates collaboration among point-cloud neighbors (regressors) directly into the data-modeling term of the loss function. Multiple kernel functions are allowed to offer robustness and rich approximation properties, while multiple matrix factors offer low-rank modeling, integrate dimensionality reduction, and streamline computations with no need of training data. Two important application domains showcase the functionality of MultiL-KRIM: time-varying-graph-signal (TVGS) recovery, and reconstruction of highly accelerated dynamic-magnetic-resonance-imaging (dMRI) data. Extensive numerical tests on real and synthetic data demonstrate MultiL-KRIM's remarkable speedups over its predecessors, and outperformance over prevalent "shallow" data-imputation techniques, with a more intuitive and explainable pipeline than deep-image-prior methods.
Large Language Models (LLMs) have seen significant use in domains such as natural language processing and computer vision. Going beyond text, image and graphics, LLMs present a significant potential for analysis of time series data, benefiting domains such as climate, IoT, healthcare, traffic, audio and finance. This survey paper provides an in-depth exploration and a detailed taxonomy of the various methodologies employed to harness the power of LLMs for time series analysis. We address the inherent challenge of bridging the gap between LLMs' original text data training and the numerical nature of time series data, and explore strategies for transferring and distilling knowledge from LLMs to numerical time series analysis. We detail various methodologies, including (1) direct prompting of LLMs, (2) time series quantization, (3) alignment techniques, (4) utilization of the vision modality as a bridging mechanism, and (5) the combination of LLMs with tools. Additionally, this survey offers a comprehensive overview of the existing multimodal time series and text datasets and delves into the challenges and future opportunities of this emerging field. We maintain an up-to-date Github repository which includes all the papers and datasets discussed in the survey.
We define and investigate the problem of $\textit{c-approximate window search}$: approximate nearest neighbor search where each point in the dataset has a numeric label, and the goal is to find nearest neighbors to queries within arbitrary label ranges. Many semantic search problems, such as image and document search with timestamp filters, or product search with cost filters, are natural examples of this problem. We propose and theoretically analyze a modular tree-based framework for transforming an index that solves the traditional c-approximate nearest neighbor problem into a data structure that solves window search. On standard nearest neighbor benchmark datasets equipped with random label values, adversarially constructed embeddings, and image search embeddings with real timestamps, we obtain up to a $75\times$ speedup over existing solutions at the same level of recall.
When enough annotated training data is available, supervised deep-learning algorithms excel at estimating human body pose and shape using a single camera. The effects of too little such data being available can be mitigated by using other information sources, such as databases of body shapes, to learn priors. Unfortunately, such sources are not always available either. We show that, in such cases, easy-to-obtain unannotated videos can be used instead to provide the required supervisory signals. Given a trained model using too little annotated data, we compute poses in consecutive frames along with the optical flow between them. We then enforce consistency between the image optical flow and the one that can be inferred from the change in pose from one frame to the next. This provides enough additional supervision to effectively refine the network weights and to perform on par with methods trained using far more annotated data.
In recent years, groundbreaking advancements in Generative Artificial Intelligence (GenAI) have triggered a transformative paradigm shift, significantly influencing various domains. In this work, we specifically explore an integrated approach, leveraging advanced techniques in GenAI and computer vision emphasizing image manipulation. The methodology unfolds through several stages, including depth estimation, the creation of inpaint masks based on depth information, the generation and replacement of backgrounds utilizing Stable Diffusion in conjunction with Latent Consistency Models (LCMs), and the subsequent replacement of clothes and application of aesthetic changes through an inpainting pipeline. Experiments conducted in this study underscore the methodology's efficacy, highlighting its potential to produce visually captivating content. The convergence of these advanced techniques allows users to input photographs of individuals and manipulate them to modify clothing and background based on specific prompts without manually input inpainting masks, effectively placing the subjects within the vast landscape of creative imagination.
We investigate a challenging task of nighttime optical flow, which suffers from weakened texture and amplified noise. These degradations weaken discriminative visual features, thus causing invalid motion feature matching. Typically, existing methods employ domain adaptation to transfer knowledge from auxiliary domain to nighttime domain in either input visual space or output motion space. However, this direct adaptation is ineffective, since there exists a large domain gap due to the intrinsic heterogeneous nature of the feature representations between auxiliary and nighttime domains. To overcome this issue, we explore a common-latent space as the intermediate bridge to reinforce the feature alignment between auxiliary and nighttime domains. In this work, we exploit two auxiliary daytime and event domains, and propose a novel common appearance-boundary adaptation framework for nighttime optical flow. In appearance adaptation, we employ the intrinsic image decomposition to embed the auxiliary daytime image and the nighttime image into a reflectance-aligned common space. We discover that motion distributions of the two reflectance maps are very similar, benefiting us to consistently transfer motion appearance knowledge from daytime to nighttime domain. In boundary adaptation, we theoretically derive the motion correlation formula between nighttime image and accumulated events within a spatiotemporal gradient-aligned common space. We figure out that the correlation of the two spatiotemporal gradient maps shares significant discrepancy, benefitting us to contrastively transfer boundary knowledge from event to nighttime domain. Moreover, appearance adaptation and boundary adaptation are complementary to each other, since they could jointly transfer global motion and local boundary knowledge to the nighttime domain.
In the field of computer vision, the persistent presence of color bias, resulting from fluctuations in real-world lighting and camera conditions, presents a substantial challenge to the robustness of models. This issue is particularly pronounced in complex wide-area surveillance scenarios, such as person re-identification and industrial dust segmentation, where models often experience a decline in performance due to overfitting on color information during training, given the presence of environmental variations. Consequently, there is a need to effectively adapt models to cope with the complexities of camera conditions. To address this challenge, this study introduces a learning strategy named Random Color Erasing, which draws inspiration from ensemble learning. This strategy selectively erases partial or complete color information in the training data without disrupting the original image structure, thereby achieving a balanced weighting of color features and other features within the neural network. This approach mitigates the risk of overfitting and enhances the model's ability to handle color variation, thereby improving its overall robustness. The approach we propose serves as an ensemble learning strategy, characterized by robust interpretability. A comprehensive analysis of this methodology is presented in this paper. Across various tasks such as person re-identification and semantic segmentation, our approach consistently improves strong baseline methods. Notably, in comparison to existing methods that prioritize color robustness, our strategy significantly enhances performance in cross-domain scenarios. The code available at \url{https://github.com/layumi/Person\_reID\_baseline\_pytorch/blob/master/random\_erasing.py} or \url{https://github.com/finger-monkey/Data-Augmentation}.
We concentrate on a novel human-centric image synthesis task, that is, given only one reference facial photograph, it is expected to generate specific individual images with diverse head positions, poses, and facial expressions in different contexts. To accomplish this goal, we argue that our generative model should be capable of the following favorable characteristics: (1) a strong visual and semantic understanding of our world and human society for basic object and human image generation. (2) generalizable identity preservation ability. (3) flexible and fine-grained head control. Recently, large pre-trained text-to-image diffusion models have shown remarkable results, serving as a powerful generative foundation. As a basis, we aim to unleash the above two capabilities of the pre-trained model. In this work, we present a new framework named CapHuman. We embrace the ``encode then learn to align" paradigm, which enables generalizable identity preservation for new individuals without cumbersome tuning at inference. CapHuman encodes identity features and then learns to align them into the latent space. Moreover, we introduce the 3D facial prior to equip our model with control over the human head in a flexible and 3D-consistent manner. Extensive qualitative and quantitative analyses demonstrate our CapHuman can produce well-identity-preserved, photo-realistic, and high-fidelity portraits with content-rich representations and various head renditions, superior to established baselines. Code and checkpoint will be released at https://github.com/VamosC/CapHuman.