머신 언러닝(Machine unlearning)은 모델을 처음부터 다시 학습(training)하지 않고도 학습된 모델(model)에서 특정 학습 데이터의 영향을 제거합니다. 언러닝 기법을 평가하려면 여러 시드(seed)에 걸쳐 학습, 언러닝, 평가를 반복해야 하므로 계산 비용이 많이 듭니다. 우리가 아는 한, 기존 이미지 분류(image classification) 언러닝 프레임워크는 단일 GPU에서 실행되므로 적절한 시간 내에 평가할 수 있는 시드 수가 제한됩니다. 본 논문에서는 이러한 단계를 여러 GPU에 분산하는 오픈소스 프레임워크인 SUPREME을 소개합니다. SUPREME은 새로운 기법, 지표, 모델, 시나리오를 추가하기 위한 레지스트리 기반 설계, 여러 가속기와 정밀도(precision) 모드를 지원하는 멀티 GPU 아키텍처, 그리고 10개의 시드에 걸쳐 전체 클래스 및 무작위 샘플 언러닝 조건에서 ResNet18 및 ViT를 사용해 Pins Face Recognition 데이터셋에 대해 진행한 시연이라는 세 가지 기여를 제공합니다. 해당 프레임워크는 https://github.com/pedroandreou/supreme-unlearning 에서 이용할 수 있습니다.
We present the first approach to automated audio captioning. We employ an encoder-decoder scheme with an alignment model in between. The input to the encoder is a sequence of log mel-band energies calculated from an audio file, while the output is a sequence of words, i.e. a caption. The encoder is a multi-layered, bi-directional gated recurrent unit (GRU) and the decoder a multi-layered GRU with a classification layer connected to the last GRU of the decoder. The classification layer and the alignment model are fully connected layers with shared weights between timesteps. The proposed method is evaluated using data drawn from a commercial sound effects library, ProSound Effects. The resulting captions were rated through metrics utilized in machine translation and image captioning fields. Results from metrics show that the proposed method can predict words appearing in the original caption, but not always correctly ordered.
This paper attempts to find the most accurate classification method among parallelepiped, minimum distance and chain methods. Moreover, this study also challenges to find the suitable combination of bands, which can lead to better results in case combinations of bands occur. After comparing these three methods, the chain method over perform the other methods with 79% overall accuracy. Hence, it is more accurate than minimum distance with 67% and parallelepiped with 65%. On the other hand, based on bands features, and also by combining several researchers' findings, a table was created which includes the main objects on the land and the suitable combination of the bands for accurately detecting of landcover objects. During this process, it was observed that band 4 (out of 7 bands of Landsat 5TM) is the band, which can be used for increasing the accuracy of the combined bands in detecting objects on the land.
This paper proposes a method to use deep neural networks as end-to-end open-set classifiers. It is based on intra-class data splitting. In open-set recognition, only samples from a limited number of known classes are available for training. During inference, an open-set classifier must reject samples from unknown classes while correctly classifying samples from known classes. The proposed method splits given data into typical and atypical normal subsets by using a closed-set classifier. This enables to model the abnormal classes by atypical normal samples. Accordingly, the open-set recognition problem is reformulated into a traditional classification problem. In addition, a closed-set regularization is proposed to guarantee a high closed-set classification performance. Intensive experiments on five well-known image datasets showed the effectiveness of the proposed method which outperformed the baselines and achieved a distinct improvement over the state-of-the-art methods.
We present an interpretable deep model for fine-grained visual recognition. At the core of our method lies the integration of region-based part discovery and attribution within a deep neural network. Our model is trained using image-level object labels, and provides an interpretation of its results via the segmentation of object parts and the identification of their contributions towards classification. To facilitate the learning of object parts without direct supervision, we explore a simple prior of the occurrence of object parts. We demonstrate that this prior, when combined with our region-based part discovery and attribution, leads to an interpretable model that remains highly accurate. Our model is evaluated on major fine-grained recognition datasets, including CUB-200, CelebA and iNaturalist. Our results compare favorably to state-of-the-art methods on classification tasks, and our method outperforms previous approaches on the localization of object parts.
Self-supervised contrastive learning has demonstrated great potential in learning visual representations. Despite their success in various downstream tasks such as image classification and object detection, self-supervised pre-training for fine-grained scenarios is not fully explored. We point out that current contrastive methods are prone to memorizing background/foreground texture and therefore have a limitation in localizing the foreground object. Analysis suggests that learning to extract discriminative texture information and localization are equally crucial for fine-grained self-supervised pre-training. Based on our findings, we introduce cross-view saliency alignment (CVSA), a contrastive learning framework that first crops and swaps saliency regions of images as a novel view generation and then guides the model to localize on foreground objects via a cross-view alignment loss. Extensive experiments on both small- and large-scale fine-grained classification benchmarks show that CVSA significantly improves the learned representation.
Predictive uncertainties in classification tasks are often a consequence of model inadequacy or insufficient training data. In popular applications, such as image processing, we are often required to scrutinise these uncertainties by meaningfully attributing them to input features. This helps to improve interpretability assessments. However, there exist few effective frameworks for this purpose. Vanilla forms of popular methods for the provision of saliency masks, such as SHAP or integrated gradients, adapt poorly to target measures of uncertainty. Thus, state-of-the-art tools instead proceed by creating counterfactual or adversarial feature vectors, and assign attributions by direct comparison to original images. In this paper, we present a novel framework that combines path integrals, counterfactual explanations and generative models, in order to procure attributions that contain few observable artefacts or noise. We evidence that this outperforms existing alternatives through quantitative evaluations with popular benchmarking methods and data sets of varying complexity.
State-of-the-art image and text classification models, such as Convolutional Neural Networks and Transformers, have long been able to classify their respective unimodal reasoning satisfactorily with accuracy close to or exceeding human accuracy. However, images embedded with text, such as hateful memes, are hard to classify using unimodal reasoning when difficult examples, such as benign confounders, are incorporated into the data set. We attempt to generate more labeled memes in addition to the Hateful Memes data set from Facebook AI, based on the framework of a winning team from the Hateful Meme Challenge. To increase the number of labeled memes, we explore semi-supervised learning using pseudo-labels for newly introduced, unlabeled memes gathered from the Memotion Dataset 7K. We find that the semi-supervised learning task on unlabeled data required human intervention and filtering and that adding a limited amount of new data yields no extra classification performance.
Understanding what deep network models capture in their learned representations is a fundamental challenge in computer vision. We present a new methodology to understanding such vision models, the Visual Concept Connectome (VCC), which discovers human interpretable concepts and their interlayer connections in a fully unsupervised manner. Our approach simultaneously reveals fine-grained concepts at a layer, connection weightings across all layers and is amendable to global analysis of network structure (e.g., branching pattern of hierarchical concept assemblies). Previous work yielded ways to extract interpretable concepts from single layers and examine their impact on classification, but did not afford multilayer concept analysis across an entire network architecture. Quantitative and qualitative empirical results show the effectiveness of VCCs in the domain of image classification. Also, we leverage VCCs for the application of failure mode debugging to reveal where mistakes arise in deep networks.
In this paper we introduce RE-tune, a novel approach for fine-tuning pre-trained Multimodal Biomedical Vision-Language models (VLMs) in Incremental Learning scenarios for multi-label chest disease diagnosis. RE-tune freezes the backbones and only trains simple adaptors on top of the Image and Text encoders of the VLM. By engineering positive and negative text prompts for diseases, we leverage the ability of Large Language Models to steer the training trajectory. We evaluate RE-tune in three realistic incremental learning scenarios: class-incremental, label-incremental, and data-incremental. Our results demonstrate that Biomedical VLMs are natural continual learners and prevent catastrophic forgetting. RE-tune not only achieves accurate multi-label classification results, but also prioritizes patient privacy and it distinguishes itself through exceptional computational efficiency, rendering it highly suitable for broad adoption in real-world healthcare settings.
Supervised object detection and semantic segmentation require object or even pixel level annotations. When there exist image level labels only, it is challenging for weakly supervised algorithms to achieve accurate predictions. The accuracy achieved by top weakly supervised algorithms is still significantly lower than their fully supervised counterparts. In this paper, we propose a novel weakly supervised curriculum learning pipeline for multi-label object recognition, detection and semantic segmentation. In this pipeline, we first obtain intermediate object localization and pixel labeling results for the training images, and then use such results to train task-specific deep networks in a fully supervised manner. The entire process consists of four stages, including object localization in the training images, filtering and fusing object instances, pixel labeling for the training images, and task-specific network training. To obtain clean object instances in the training images, we propose a novel algorithm for filtering, fusing and classifying object instances collected from multiple solution mechanisms. In this algorithm, we incorporate both metric learning and density-based clustering to filter detected object instances. Experiments show that our weakly supervised pipeline achieves state-of-the-art results in multi-label image classification as well as weakly supervised object detection and very competitive results in weakly supervised semantic segmentation on MS-COCO, PASCAL VOC 2007 and PASCAL VOC 2012.
Purpose of review: This paper presents a review of the current state of the art in remote sensing based monitoring of forest disturbances and forest degradation from optical Earth Observation data. Part one comprises an overview of currently available optical remote sensing sensors, which can be used for forest disturbance and degradation mapping. Part two reviews the two main categories of existing approaches: classical image-to-image change detection and time series analysis. Recent findings: With the launch of the Sentinel-2a satellite and available Landsat imagery, time series analysis has become the most promising but also most demanding category of degradation mapping approaches. Four time series classification methods are distinguished. The methods are explained and their benefits and drawbacks are discussed. A separate chapter presents a number of recent forest degradation mapping studies for two different ecosystems: temperate forests with a geographical focus on Europe and tropical forests with a geographical focus on Africa. Summary: The review revealed that a wide variety of methods for the detection of forest degradation is already available. Today, the main challenge is to transfer these approaches to high resolution time series data from multiple sensors. Future research should also focus on the classification of disturbance types and the development of robust up-scalable methods to enable near real time disturbance mapping in support of operational reactive measures.
In recent years, people from all over the world are suffering from one of the most severe diseases in history, known as Coronavirus disease 2019, COVID-19 for short. When the virus reaches the lungs, it has a higher probability to cause lung pneumonia and sepsis. X-ray image is a powerful tool in identifying the typical features of the infection for COVID-19 patients. The radiologists and pathologists observe that ground-glass opacity appears in the chest X-ray for infected patient \cite{cozzi2021ground}, and it could be used as one of the criteria during the diagnosis process. In the past few years, deep learning has proven to be one of the most powerful methods in the field of image classification. Due to significant differences in Chest X-Ray between normal and infected people \cite{rousan2020chest}, deep models could be used to identify the presence of the disease given a patient's Chest X-Ray. Many deep models are complex, and it evolves with lots of input parameters. Designers sometimes struggle with the tuning process for deep models, especially when they build up the model from scratch. Genetic Algorithm, inspired by the biological evolution process, plays a key role in solving such complex problems. In this paper, I proposed a genetic-based approach to optimize the Convolutional Neural Network(CNN) for the Chest X-Ray classification task.
Histopathological image analysis is the gold standard to diagnose cancer. Carcinoma is a subtype of cancer that constitutes more than 80% of all cancer cases. Squamous cell carcinoma and adenocarcinoma are two major subtypes of carcinoma, diagnosed by microscopic study of biopsy slides. However, manual microscopic evaluation is a subjective and time-consuming process. Many researchers have reported methods to automate carcinoma detection and classification. The increasing use of artificial intelligence (AI) in the automation of carcinoma diagnosis also reveals a significant rise in the use of deep network models. In this systematic literature review, we present a comprehensive review of the state-of-the-art approaches reported in carcinoma diagnosis using histopathological images. Studies are selected from well-known databases with strict inclusion/exclusion criteria. We have categorized the articles and recapitulated their methods based on specific organs of carcinoma origin. Further, we have summarized pertinent literature on AI methods, highlighted critical challenges and limitations, and provided insights on future research direction in automated carcinoma diagnosis. Out of 101 articles selected, most of the studies experimented on private datasets with varied image sizes, obtaining accuracy between 63% and 100%. Overall, this review highlights the need for a generalized AI-based carcinoma diagnostic system. Additionally, it is desirable to have accountable approaches to extract microscopic features from images of multiple magnifications that should mimic pathologists' evaluations.
Determining the type of kidney stones allows urologists to prescribe a treatment to avoid recurrence of renal lithiasis. An automated in-vivo image-based classification method would be an important step towards an immediate identification of the kidney stone type required as a first phase of the diagnosis. In the literature it was shown on ex-vivo data (i.e., in very controlled scene and image acquisition conditions) that an automated kidney stone classification is indeed feasible. This pilot study compares the kidney stone recognition performances of six shallow machine learning methods and three deep-learning architectures which were tested with in-vivo images of the four most frequent urinary calculi types acquired with an endoscope during standard ureteroscopies. This contribution details the database construction and the design of the tested kidney stones classifiers. Even if the best results were obtained by the Inception v3 architecture (weighted precision, recall and F1-score of 0.97, 0.98 and 0.97, respectively), it is also shown that choosing an appropriate colour space and texture features allows a shallow machine learning method to approach closely the performances of the most promising deep-learning methods (the XGBoost classifier led to weighted precision, recall and F1-score values of 0.96). This paper is the first one that explores the most discriminant features to be extracted from images acquired during ureteroscopies.
Marine scientists use remote underwater video recording to survey fish species in their natural habitats. This helps them understand and predict how fish respond to climate change, habitat degradation, and fishing pressure. This information is essential for developing sustainable fisheries for human consumption, and for preserving the environment. However, the enormous volume of collected videos makes extracting useful information a daunting and time-consuming task for a human. A promising method to address this problem is the cutting-edge Deep Learning (DL) technology.DL can help marine scientists parse large volumes of video promptly and efficiently, unlocking niche information that cannot be obtained using conventional manual monitoring methods. In this paper, we provide an overview of the key concepts of DL, while presenting a survey of literature on fish habitat monitoring with a focus on underwater fish classification. We also discuss the main challenges faced when developing DL for underwater image processing and propose approaches to address them. Finally, we provide insights into the marine habitat monitoring research domain and shed light on what the future of DL for underwater image processing may hold. This paper aims to inform a wide range of readers from marine scientists who would like to apply DL in their research to computer scientists who would like to survey state-of-the-art DL-based underwater fish habitat monitoring literature.
There has been significant recent progress in training differentially private (DP) models which achieve accuracy that approaches the best non-private models. These DP models are typically pretrained on large public datasets and then fine-tuned on private downstream datasets that are relatively large and similar in distribution to the pretraining data. However, in many applications including personalization and federated learning, it is crucial to perform well (i) in the few-shot setting, as obtaining large amounts of labeled data may be problematic; and (ii) on datasets from a wide variety of domains for use in various specialist settings. To understand under which conditions few-shot DP can be effective, we perform an exhaustive set of experiments that reveals how the accuracy and vulnerability to attack of few-shot DP image classification models are affected as the number of shots per class, privacy level, model architecture, downstream dataset, and subset of learnable parameters in the model vary. We show that to achieve DP accuracy on par with non-private models, the shots per class must be increased as the privacy level increases. We also show that learning parameter-efficient FiLM adapters under DP is competitive with learning just the final classifier layer or learning all of the network parameters. Finally, we evaluate DP federated learning systems and establish state-of-the-art performance on the challenging FLAIR benchmark.
Deep learning models have proven to be highly successful. Yet, their over-parameterization gives rise to model multiplicity, a phenomenon in which multiple models achieve similar performance but exhibit distinct underlying behaviours. This multiplicity presents a significant challenge and necessitates additional specifications in model selection to prevent unexpected failures during deployment. While prior studies have examined these concerns, they focus on individual metrics in isolation, making it difficult to obtain a comprehensive view of multiplicity in trustworthy machine learning. Our work stands out by offering a one-stop empirical benchmark of multiplicity across various dimensions of model design and its impact on a diverse set of trustworthy metrics. In this work, we establish a consistent language for studying model multiplicity by translating several trustworthy metrics into accuracy under appropriate interventions. We also develop a framework, which we call multiplicity sheets, to benchmark multiplicity in various scenarios. We demonstrate the advantages of our setup through a case study in image classification and provide actionable insights into the impact and trends of different hyperparameters on model multiplicity. Finally, we show that multiplicity persists in deep learning models even after enforcing additional specifications during model selection, highlighting the severity of over-parameterization. The concerns of under-specification thus remain, and we seek to promote a more comprehensive discussion of multiplicity in trustworthy machine learning.
With the rapid development of imaging sensor technology in the field of remote sensing, multi-modal remote sensing data fusion has emerged as a crucial research direction for land cover classification tasks. While diffusion models have made great progress in generative models and image classification tasks, existing models primarily focus on single-modality and single-client control, that is, the diffusion process is driven by a single modal in a single computing node. To facilitate the secure fusion of heterogeneous data from clients, it is necessary to enable distributed multi-modal control, such as merging the hyperspectral data of organization A and the LiDAR data of organization B privately on each base station client. In this study, we propose a multi-modal collaborative diffusion federated learning framework called FedDiff. Our framework establishes a dual-branch diffusion model feature extraction setup, where the two modal data are inputted into separate branches of the encoder. Our key insight is that diffusion models driven by different modalities are inherently complementary in terms of potential denoising steps on which bilateral connections can be built. Considering the challenge of private and efficient communication between multiple clients, we embed the diffusion model into the federated learning communication structure, and introduce a lightweight communication module. Qualitative and quantitative experiments validate the superiority of our framework in terms of image quality and conditional consistency.
Continual learning is a challenging problem in machine learning, especially for image classification tasks with imbalanced datasets. It becomes even more challenging when it involves learning new classes incrementally. One method for incremental class learning, addressing dataset imbalance, is rehearsal using previously stored data. In rehearsal-based methods, access to previous data is required for either training the classifier or the generator, but it may not be feasible due to storage, legal, or data access constraints. Although there are many rehearsal-free alternatives for class incremental learning, such as parameter or loss regularization, knowledge distillation, and dynamic architectures, they do not consistently achieve good results, especially on imbalanced data. This paper proposes a new approach called Data-Free Generative Replay (DFGR) for class incremental learning, where the generator is trained without access to real data. In addition, DFGR also addresses dataset imbalance in continual learning of an image classifier. Instead of using training data, DFGR trains a generator using mean and variance statistics of batch-norm and feature maps derived from a pre-trained classification model. The results of our experiments demonstrate that DFGR performs significantly better than other data-free methods and reveal the performance impact of specific parameter settings. DFGR achieves up to 88.5% and 46.6% accuracy on MNIST and FashionMNIST datasets, respectively. Our code is available at https://github.com/2younis/DFGR