Existing visual tracking methods usually localize a target object with a bounding box, in which the performance of the foreground object trackers or detectors is often affected by the inclusion of background clutter. To handle this problem, we learn a patch-based graph representation for visual tracking. The tracked object is modeled by with a graph by taking a set of non-overlapping image patches as nodes, in which the weight of each node indicates how likely it belongs to the foreground and edges are weighted for indicating the appearance compatibility of two neighboring nodes. This graph is dynamically learned and applied in object tracking and model updating. During the tracking process, the proposed algorithm performs three main steps in each frame. First, the graph is initialized by assigning binary weights of some image patches to indicate the object and background patches according to the predicted bounding box. Second, the graph is optimized to refine the patch weights by using a novel alternating direction method of multipliers. Third, the object feature representation is updated by imposing the weights of patches on the extracted image features. The object location is predicted by maximizing the classification score in the structured support vector machine. Extensive experiments show that the proposed tracking algorithm performs well against the state-of-the-art methods on large-scale benchmark datasets.
We present a simple and effective architecture for fine-grained visual recognition called Bilinear Convolutional Neural Networks (B-CNNs). These networks represent an image as a pooled outer product of features derived from two CNNs and capture localized feature interactions in a translationally invariant manner. B-CNNs belong to the class of orderless texture representations but unlike prior work they can be trained in an end-to-end manner. Our most accurate model obtains 84.1%, 79.4%, 86.9% and 91.3% per-image accuracy on the Caltech-UCSD birds [67], NABirds [64], FGVC aircraft [42], and Stanford cars [33] dataset respectively and runs at 30 frames-per-second on a NVIDIA Titan X GPU. We then present a systematic analysis of these networks and show that (1) the bilinear features are highly redundant and can be reduced by an order of magnitude in size without significant loss in accuracy, (2) are also effective for other image classification tasks such as texture and scene recognition, and (3) can be trained from scratch on the ImageNet dataset offering consistent improvements over the baseline architecture. Finally, we present visualizations of these models on various datasets using top activations of neural units and gradient-based inversion techniques. The source code for the complete system is available at http://vis-www.cs.umass.edu/bcnn.
Providing huge amounts of data is a fundamental demand when dealing with Deep Neural Networks (DNNs). Employing these algorithms to solve computer vision problems resulted in the advent of various image datasets to feed the most common visual imagery deep structures, known as Convolutional Neural Networks (CNNs). In this regard, some datasets can be found that contain hundreds or even thousands of images for license plate detection and optical character recognition purposes. However, no publicly available image dataset provides such data for the recognition of Farsi characters used in car license plates. The gap has to be filled due to the numerous advantages of developing accurate deep learning-based systems for law enforcement and surveillance purposes. This paper introduces a large-scale dataset that includes images of numbers and characters used in Iranian car license plates. The dataset, named Iranis, contains more than 83,000 images of Farsi numbers and letters collected from real-world license plate images captured by various cameras. The variety of instances in terms of camera shooting angle, illumination, resolution, and contrast make the dataset a proper choice for training DNNs. Dataset images are manually annotated for object detection and image classification. Finally, and to build a baseline for Farsi character recognition, the paper provides a performance analysis using a YOLO v.3 object detector.
Image-level weakly supervised semantic segmentation (WSSS) is a fundamental yet challenging computer vision task facilitating scene understanding and automatic driving. Most existing methods resort to classification-based Class Activation Maps (CAMs) to play as the initial pseudo labels, which tend to focus on the discriminative image regions and lack customized characteristics for the segmentation task. To alleviate this issue, we propose a novel activation modulation and recalibration (AMR) scheme, which leverages a spotlight branch and a compensation branch to obtain weighted CAMs that can provide recalibration supervision and task-specific concepts. Specifically, an attention modulation module (AMM) is employed to rearrange the distribution of feature importance from the channel-spatial sequential perspective, which helps to explicitly model channel-wise interdependencies and spatial encodings to adaptively modulate segmentation-oriented activation responses. Furthermore, we introduce a cross pseudo supervision for dual branches, which can be regarded as a semantic similar regularization to mutually refine two branches. Extensive experiments show that AMR establishes a new state-of-the-art performance on the PASCAL VOC 2012 dataset, surpassing not only current methods trained with the image-level of supervision but also some methods relying on stronger supervision, such as saliency label. Experiments also reveal that our scheme is plug-and-play and can be incorporated with other approaches to boost their performance.
We propose a Multi-level Second-order (MlSo) few-shot learning network for supervised or unsupervised few-shot image classification and few-shot action recognition. We leverage so-called power-normalized second-order base learner streams combined with features that express multiple levels of visual abstraction, and we use self-supervised discriminating mechanisms. As Second-order Pooling (SoP) is popular in image recognition, we employ its basic element-wise variant in our pipeline. The goal of multi-level feature design is to extract feature representations at different layer-wise levels of CNN, realizing several levels of visual abstraction to achieve robust few-shot learning. As SoP can handle convolutional feature maps of varying spatial sizes, we also introduce image inputs at multiple spatial scales into MlSo. To exploit the discriminative information from multi-level and multi-scale features, we develop a Feature Matching (FM) module that reweights their respective branches. We also introduce a self-supervised step, which is a discriminator of the spatial level and the scale of abstraction. Our pipeline is trained in an end-to-end manner. With a simple architecture, we demonstrate respectable results on standard datasets such as Omniglot, mini-ImageNet, tiered-ImageNet, Open MIC, fine-grained datasets such as CUB Birds, Stanford Dogs and Cars, and action recognition datasets such as HMDB51, UCF101, and mini-MIT.
Image segmentation is a fundamental task in computer vision. Data annotation for training supervised methods can be labor-intensive, motivating unsupervised methods. Current approaches often rely on extracting deep features from pre-trained networks to construct a graph, and classical clustering methods like k-means and normalized-cuts are then applied as a post-processing step. However, this approach reduces the high-dimensional information encoded in the features to pair-wise scalar affinities. To address this limitation, this study introduces a lightweight Graph Neural Network (GNN) to replace classical clustering methods while optimizing for the same clustering objective function. Unlike existing methods, our GNN takes both the pair-wise affinities between local image features and the raw features as input. This direct connection between the raw features and the clustering objective enables us to implicitly perform classification of the clusters between different graphs, resulting in part semantic segmentation without the need for additional post-processing steps. We demonstrate how classical clustering objectives can be formulated as self-supervised loss functions for training an image segmentation GNN. Furthermore, we employ the Correlation-Clustering (CC) objective to perform clustering without defining the number of clusters, allowing for k-less clustering. We apply the proposed method for object localization, segmentation, and semantic part segmentation tasks, surpassing state-of-the-art performance on multiple benchmarks.
Graph neural networks (GNNs) present a promising alternative to CNNs and transformers in certain image processing applications due to their parameter-efficiency in modeling spatial relationships. Currently, a major area of research involves the converting non-graph input data for GNN-based models, notably in scenarios where the data originates from images. One approach involves converting images into nodes by identifying significant keypoints within them. Super-Retina, a semi-supervised technique, has been utilized for detecting keypoints in retinal images. However, its limitations lie in the dependency on a small initial set of ground truth keypoints, which is progressively expanded to detect more keypoints. Having encountered difficulties in detecting consistent initial keypoints in brain images using SIFT and LoFTR, we proposed a new approach: radiomic feature-based keypoint detection. Demonstrating the anatomical significance of the detected keypoints was achieved by showcasing their efficacy in improving registration processes guided by these keypoints. Subsequently, these keypoints were employed as the ground truth for the keypoint detection method (LK-SuperRetina). Furthermore, the study showcases the application of GNNs in image matching, highlighting their superior performance in terms of both the number of good matches and confidence scores. This research sets the stage for expanding GNN applications into various other applications, including but not limited to image classification, segmentation, and registration.
Convolutional Neural Networks (CNNs) have advanced significantly in visual representation learning and recognition. However, they face notable challenges in performance and computational efficiency when dealing with real-world, multi-scale image inputs. Conventional methods rescale all input images into a fixed size, wherein a larger fixed size favors performance but rescaling small size images to a larger size incurs digitization noise and increased computation cost. In this work, we carry out a comprehensive, layer-wise investigation of CNN models in response to scale variation, based on Centered Kernel Alignment (CKA) analysis. The observations reveal lower layers are more sensitive to input image scale variations than high-level layers. Inspired by this insight, we propose Multi-scale Unified Network (MUSN) consisting of multi-scale subnets, a unified network, and scale-invariant constraint. Our method divides the shallow layers into multi-scale subnets to enable feature extraction from multi-scale inputs, and the low-level features are unified in deep layers for extracting high-level semantic features. A scale-invariant constraint is posed to maintain feature consistency across different scales. Extensive experiments on ImageNet and other scale-diverse datasets, demonstrate that MSUN achieves significant improvements in both model performance and computational efficiency. Particularly, MSUN yields an accuracy increase up to 44.53% and diminishes FLOPs by 7.01-16.13% in multi-scale scenarios.
The preservation of aquatic biodiversity is critical in mitigating the effects of climate change. Aquatic scene understanding plays a pivotal role in aiding marine scientists in their decision-making processes. In this paper, we introduce AquaticCLIP, a novel contrastive language-image pre-training model tailored for aquatic scene understanding. AquaticCLIP presents a new unsupervised learning framework that aligns images and texts in aquatic environments, enabling tasks such as segmentation, classification, detection, and object counting. By leveraging our large-scale underwater image-text paired dataset without the need for ground-truth annotations, our model enriches existing vision-language models in the aquatic domain. For this purpose, we construct a 2 million underwater image-text paired dataset using heterogeneous resources, including YouTube, Netflix, NatGeo, etc. To fine-tune AquaticCLIP, we propose a prompt-guided vision encoder that progressively aggregates patch features via learnable prompts, while a vision-guided mechanism enhances the language encoder by incorporating visual context. The model is optimized through a contrastive pretraining loss to align visual and textual modalities. AquaticCLIP achieves notable performance improvements in zero-shot settings across multiple underwater computer vision tasks, outperforming existing methods in both robustness and interpretability. Our model sets a new benchmark for vision-language applications in underwater environments. The code and dataset for AquaticCLIP are publicly available on GitHub at xxx.
다중 모달 융합은 모달리티 간 정보 교환뿐 아니라 어떤 표현이 상호작용에 참여해야 하는지, 그리고 상보적인 교차 모달 관점을 어떻게 결합해야 하는지를 결정하는 것도 요구한다. 본 논문에서는 토큰 단위 유도, 양방향 교차 어텐션, 샘플별 융합을 통합한 융합 프레임워크인 적응형 게이팅 기반 유도 공동 어텐션(Guided Co-Attention with Adaptive Gating)을 소개한다. 제안 방법은 두 가지 방향성 표현을 구성한다. 하나는 텍스트 토큰이 시각 특징을 질의하도록 하여 얻는 이미지 조건부 텍스트 표현이고, 다른 하나는 반대 방향으로 형성되는 텍스트 조건부 이미지 표현이다. 토큰 단위 유도는 교차 어텐션 이전에 표현을 조절하며, 적응형 게이트는 각 샘플에 대해 결과로 얻어진 두 방향성 표현의 상대적 기여도를 학습한다. 우리는 흉부 X선 영상과 판독문 쌍을 이미지-텍스트 테스트베드로 사용하여 이진(정상 대 비정상) 분류에서 이 접근법을 평가한다. 두 교차 어텐션 방향은 서로 다른 성능 특성을 보이며, 질의와 소스 모달리티를 뒤바꾸어도 서로 교환 가능한 표현이 만들어지지 않음을 보여 준다. 절제 실험은 양방향 상호작용, 적응형 게이팅, 토큰 유도가 상보적인 이점을 제공함을 추가로 보여 준다. 제안 모델은 정확도 96.05%, F1 점수 96.10%, AUC 98.37%, Sens@95Spec 95.60%, Spec@95Sens 96.51%를 달성하여 평가된 모든 지표에서 가장 높은 평균 성능을 기록하였다. 이러한 결과는 다중 모달 표현을 구성할 때 상호작용 방향, 토큰 관련성, 샘플 의존적 통합을 명시적으로 모델링하는 것을 뒷받침한다.
대조 목표로 학습된 CLIP류 비전-언어 모델(VLM)은 강한 전역 이미지-텍스트 표현을 학습하지만, 유클리드 임베딩과 전역 풀링은 부분-전체 및 부모-자식 관계와 같은 관계 구조를 인코딩하지 못한다. 하이퍼볼릭 VLM은 함의(entailment) 기반 목표로 이 공백을 메우며, 텍스트 조건 변형은 문장 및 구 수준 쿼리를 통해 세밀한 정렬을 개선한다. 그러나 이 두 연구 계열은 여전히 분리되어 있다. 하이퍼볼릭 VLM은 정적 이미지 및 영역 특징을 사용하는 반면, 쿼리 조건 방법은 계층적 기하 구조가 부족하다. 본 논문은 전역·지역·전역-지역 대조 학습을 쿼리 조건 시각 풀링과 결합한 계층 조건 하이퍼볼릭 VLM인 Hyper3-CLIP을 제안한다. 모델 학습을 위해 전체 캡션, 문장 조각, 지역화된 부분 설명, 추출된 구로 이루어진 경량 쿼리 계층을 텍스트에서 구성한다. 각 쿼리는 시각 패치 풀링을 조건화하며,由此 얻은 표현은 이미지-텍스트, 전체-부분, 부모-자식 함의 손실을 지원한다. 쿼리 조건 풀링은 학습 시에만 활성화된다. Hyper3-CLIP은 COCO와 Flickr에서 R@5·R@10 검색 성능을, VOC와 COCO에서 다중 레이블 분류를 개선하면서 계층 지표에서도 경쟁력을 유지한다. 또한 고정 프롬프트 체계 하에서 제로샷 프롬프트 민감도를 점검하고, 학습에 사용된 지역화 GRIT 부분 예산의 영향을 분석한다. 코드는 https://github.com/Hyper3Labs/hyper3-clip 에서 제공된다.
Convolutional Neural Networks (CNNs) have recently led to incredible breakthroughs on a variety of pattern recognition problems. Banks of finite impulse response filters are learned on a hierarchy of layers, each contributing more abstract information than the previous layer. The simplicity and elegance of the convolutional filtering process makes them perfect for structured problems such as image, video, or voice, where vertices are homogeneous in the sense of number, location, and strength of neighbors. The vast majority of classification problems, for example in the pharmaceutical, homeland security, and financial domains are unstructured. As these problems are formulated into unstructured graphs, the heterogeneity of these problems, such as number of vertices, number of connections per vertex, and edge strength, cannot be tackled with standard convolutional techniques. We propose a novel neural learning framework that is capable of handling both homogeneous and heterogeneous data, while retaining the benefits of traditional CNN successes. Recently, researchers have proposed variations of CNNs that can handle graph data. In an effort to create learnable filter banks of graphs, these methods either induce constraints on the data or require preprocessing. As opposed to spectral methods, our framework, which we term Graph-CNNs, defines filters as polynomials of functions of the graph adjacency matrix. Graph-CNNs can handle both heterogeneous and homogeneous graph data, including graphs having entirely different vertex or edge sets. We perform experiments to validate the applicability of Graph-CNNs to a variety of structured and unstructured classification problems and demonstrate state-of-the-art results on document and molecule classification problems.
Robust automated organ segmentation is a prerequisite for computer-aided diagnosis (CAD), quantitative imaging analysis and surgical assistance. For high-variability organs such as the pancreas, previous approaches report undesirably low accuracies. We present a bottom-up approach for pancreas segmentation in abdominal CT scans that is based on a hierarchy of information propagation by classifying image patches at different resolutions; and cascading superpixels. There are four stages: 1) decomposing CT slice images as a set of disjoint boundary-preserving superpixels; 2) computing pancreas class probability maps via dense patch labeling; 3) classifying superpixels by pooling both intensity and probability features to form empirical statistics in cascaded random forest frameworks; and 4) simple connectivity based post-processing. The dense image patch labeling are conducted by: efficient random forest classifier on image histogram, location and texture features; and more expensive (but with better specificity) deep convolutional neural network classification on larger image windows (with more spatial contexts). Evaluation of the approach is performed on a database of 80 manually segmented CT volumes in six-fold cross-validation (CV). Our achieved results are comparable, or better than the state-of-the-art methods (evaluated by "leave-one-patient-out"), with Dice 70.7% and Jaccard 57.9%. The computational efficiency has been drastically improved in the order of 6~8 minutes, comparing with others of ~10 hours per case. Finally, we implement a multi-atlas label fusion (MALF) approach for pancreas segmentation using the same datasets. Under six-fold CV, our bottom-up segmentation method significantly outperforms its MALF counterpart: (70.7 +/- 13.0%) versus (52.5 +/- 20.8%) in Dice. Deep CNN patch labeling confidences offer more numerical stability, reflected by smaller standard deviations.
Learning high quality class representations from few examples is a key problem in metric-learning approaches to few-shot learning. To accomplish this, we introduce a novel architecture where class representations are conditioned for each few-shot trial based on a target image. We also deviate from traditional metric-learning approaches by training a network to perform comparisons between classes rather than relying on a static metric comparison. This allows the network to decide what aspects of each class are important for the comparison at hand. We find that this flexible architecture works well in practice, achieving state-of-the-art performance on the Caltech-UCSD birds fine-grained classification task.
We introduce a general method of performing Residual Network inference and learning in the JPEG transform domain that allows the network to consume compressed images as input. Our formulation leverages the linearity of the JPEG transform to redefine convolution and batch normalization with a tune-able numerical approximation for ReLu. The result is mathematically equivalent to the spatial domain network up to the ReLu approximation accuracy. A formulation for image classification and a model conversion algorithm for spatial domain networks are given as examples of the method. We show that the sparsity of the JPEG format allows for faster processing of images with little to no penalty in the network accuracy.
This report to our stage 2 submission to the NeurIPS 2019 disentanglement challenge presents a simple image preprocessing method for learning disentangled latent factors. We propose to train a variational autoencoder on regionally aggregated feature maps obtained from networks pretrained on the ImageNet database, utilizing the implicit inductive bias contained in those features for disentanglement. This bias can be further enhanced by explicitly fine-tuning the feature maps on auxiliary tasks useful for the challenge, such as angle, position estimation, or color classification. Our approach achieved the 2nd place in stage 2 of the challenge. Code is available at https://github.com/mseitzer/neurips2019-disentanglement-challenge.
Neural networks have led to major improvements in image classification but suffer from being non-robust to adversarial changes, unreliable uncertainty estimates on out-distribution samples and their inscrutable black-box decisions. In this work we propose RATIO, a training procedure for Robustness via Adversarial Training on In- and Out-distribution, which leads to robust models with reliable and robust confidence estimates on the out-distribution. RATIO has similar generative properties to adversarial training so that visual counterfactuals produce class specific features. While adversarial training comes at the price of lower clean accuracy, RATIO achieves state-of-the-art $l_2$-adversarial robustness on CIFAR10 and maintains better clean accuracy.
ABIDE is the largest open-source autism spectrum disorder database with both fMRI data and full phenotype description. These data were extensively studied based on functional connectivity analysis as well as with deep learning on raw data, with top models accuracy close to 75\% for separate scanning sites. Yet there is still a problem of models transferability between different scanning sites within ABIDE. In the current paper, we for the first time perform domain adaptation for brain pathology classification problem on raw neuroimaging data. We use 3D convolutional autoencoders to build the domain irrelevant latent space image representation and demonstrate this method to outperform existing approaches on ABIDE data.
In this paper, we propose a new method for determining shared features of and measuring the distance between data sets or point clouds. Our approach uses the joint factorization of two data matrices $X_1,X_2$ into non-negative matrices $X_1 = AS_1, X_2 = AS_2$ to derive a similarity measure that determines how well the shared basis $A$ approximates $X_1, X_2$. We also propose a point cloud distance measure built upon this method and the learned factorization. Our method reveals structural differences in both image and text data. Potential applications include classification, detecting plagiarism or other manipulation, data denoising, and transfer learning.
This study explores a workflow for automated segmentation of lesions in FDG and PSMA PET/CT images. Due to the substantial differences in image characteristics between FDG and PSMA, specialized preprocessing steps are required. Utilizing YOLOv8 for data classification, the FDG and PSMA images are preprocessed separately before feeding them into the segmentation models, aiming to improve lesion segmentation accuracy. The study focuses on evaluating the performance of automated segmentation workflow for multitracer PET images. The findings are expected to provide critical insights for enhancing diagnostic workflows and patient-specific treatment plans. Our code will be open-sourced and available at https://github.com/jiayiliu-pku/AP2024.