We introduce EgoSchema, a very long-form video question-answering dataset, and benchmark to evaluate long video understanding capabilities of modern vision and language systems. Derived from Ego4D, EgoSchema consists of over 5000 human curated multiple choice question answer pairs, spanning over 250 hours of real video data, covering a very broad range of natural human activity and behavior. For each question, EgoSchema requires the correct answer to be selected between five given options based on a three-minute-long video clip. While some prior works have proposed video datasets with long clip lengths, we posit that merely the length of the video clip does not truly capture the temporal difficulty of the video task that is being considered. To remedy this, we introduce temporal certificate sets, a general notion for capturing the intrinsic temporal understanding length associated with a broad range of video understanding tasks & datasets. Based on this metric, we find EgoSchema to have intrinsic temporal lengths over 5.7x longer than the second closest dataset and 10x to 100x longer than any other video understanding dataset. Further, our evaluation of several current state-of-the-art video and language models shows them to be severely lacking in long-term video understanding capabilities. Even models with several billions of parameters achieve QA accuracy less than 33% (random is 20%) on the EgoSchema multi-choice question answering task, while humans achieve about 76% accuracy. We posit that \name{}{}, with its long intrinsic temporal structures and diverse complexity, would serve as a valuable evaluation probe for developing effective long-term video understanding systems in the future. Data and Zero-shot model evaluation code are open-sourced for both public and commercial use under the Ego4D license at http://egoschema.github.io

We introduce a generalized information criterion that contains other well-known information criteria, such as Bayesian information Criterion (BIC) and Akaike information criterion (AIC), as special cases. Furthermore, the proposed spectral information criterion (SIC) is also more general than the other information criteria, e.g., since the knowledge of a likelihood function is not strictly required. SIC extracts geometric features of the error curve and, as a consequence, it can be considered an automatic elbow detector. SIC provides a subset of all possible models, with a cardinality that often is much smaller than the total number of possible models. The elements of this subset are elbows of the error curve. A practical rule for selecting a unique model within the sets of elbows is suggested as well. Theoretical invariance properties of SIC are analyzed. Moreover, we test SIC in ideal scenarios where provides always the optimal expected results. We also test SIC in several numerical experiments: some involving synthetic data, and two experiments involving real datasets. They are all real-world applications such as clustering, variable selection, or polynomial order selection, to name a few. The results show the benefits of the proposed scheme. Matlab code related to the experiments is also provided. Possible future research lines are finally discussed.

For some classification scenarios, it is desirable to use only those classification instances that a trained model associates with a high certainty. To obtain such high-certainty instances, previous work has proposed accuracy-reject curves. Reject curves allow to evaluate and compare the performance of different certainty measures over a range of thresholds for accepting or rejecting classifications. However, the accuracy may not be the most suited evaluation metric for all applications, and instead precision or recall may be preferable. This is the case, for example, for data with imbalanced class distributions. We therefore propose reject curves that evaluate precision and recall, the recall-reject curve and the precision-reject curve. Using prototype-based classifiers from learning vector quantization, we first validate the proposed curves on artificial benchmark data against the accuracy reject curve as a baseline. We then show on imbalanced benchmarks and medical, real-world data that for these scenarios, the proposed precision- and recall-curves yield more accurate insights into classifier performance than accuracy reject curves.

We test similarity-based word alignment models (SimAlign and awesome-align) in combination with word embeddings from mBERT and XLM-R on parallel sentences in German and Romansh. Since Romansh is an unseen language, we are dealing with a zero-shot setting. Using embeddings from mBERT, both models reach an alignment error rate of 0.22, which outperforms fast_align, a statistical model, and is on par with similarity-based word alignment for seen languages. We interpret these results as evidence that mBERT contains information that can be meaningful and applicable to Romansh. To evaluate performance, we also present a new trilingual corpus, which we call the DERMIT (DE-RM-IT) corpus, containing press releases made by the Canton of Grisons in German, Romansh and Italian in the past 25 years. The corpus contains 4 547 parallel documents and approximately 100 000 sentence pairs in each language combination. We additionally present a gold standard for German-Romansh word alignment. The data is available at https://github.com/eyldlv/DERMIT-Corpus.

Malware detection has become a major concern due to the increasing number and complexity of malware. Traditional detection methods based on signatures and heuristics are used for malware detection, but unfortunately, they suffer from poor generalization to unknown attacks and can be easily circumvented using obfuscation techniques. In recent years, Machine Learning (ML) and notably Deep Learning (DL) achieved impressive results in malware detection by learning useful representations from data and have become a solution preferred over traditional methods. More recently, the application of such techniques on graph-structured data has achieved state-of-the-art performance in various domains and demonstrates promising results in learning more robust representations from malware. Yet, no literature review focusing on graph-based deep learning for malware detection exists. In this survey, we provide an in-depth literature review to summarize and unify existing works under the common approaches and architectures. We notably demonstrate that Graph Neural Networks (GNNs) reach competitive results in learning robust embeddings from malware represented as expressive graph structures, leading to an efficient detection by downstream classifiers. This paper also reviews adversarial attacks that are utilized to fool graph-based detection methods. Challenges and future research directions are discussed at the end of the paper.

Phytoplankton parasites are largely understudied microbial components with a potentially significant ecological impact on phytoplankton bloom dynamics. To better understand their impact, we need improved detection methods to integrate phytoplankton parasite interactions in monitoring aquatic ecosystems. Automated imaging devices usually produce high amount of phytoplankton image data, while the occurrence of anomalous phytoplankton data is rare. Thus, we propose an unsupervised anomaly detection system based on the similarity of the original and autoencoder-reconstructed samples. With this approach, we were able to reach an overall F1 score of 0.75 in nine phytoplankton species, which could be further improved by species-specific fine-tuning. The proposed unsupervised approach was further compared with the supervised Faster R-CNN based object detector. With this supervised approach and the model trained on plankton species and anomalies, we were able to reach the highest F1 score of 0.86. However, the unsupervised approach is expected to be more universal as it can detect also unknown anomalies and it does not require any annotated anomalous data that may not be always available in sufficient quantities. Although other studies have dealt with plankton anomaly detection in terms of non-plankton particles, or air bubble detection, our paper is according to our best knowledge the first one which focuses on automated anomaly detection considering putative phytoplankton parasites or infections.

An increasing number of public datasets have shown a marked impact on automated organ segmentation and tumor detection. However, due to the small size and partially labeled problem of each dataset, as well as a limited investigation of diverse types of tumors, the resulting models are often limited to segmenting specific organs/tumors and ignore the semantics of anatomical structures, nor can they be extended to novel domains. To address these issues, we propose the CLIP-Driven Universal Model, which incorporates text embedding learned from Contrastive Language-Image Pre-training (CLIP) to segmentation models. This CLIP-based label encoding captures anatomical relationships, enabling the model to learn a structured feature embedding and segment 25 organs and 6 types of tumors. The proposed model is developed from an assembly of 14 datasets, using a total of 3,410 CT scans for training and then evaluated on 6,162 external CT scans from 3 additional datasets. We rank first on the Medical Segmentation Decathlon (MSD) public leaderboard and achieve state-of-the-art results on Beyond The Cranial Vault (BTCV). Additionally, the Universal Model is computationally more efficient (6x faster) compared with dataset-specific models, generalized better to CT scans from varying sites, and shows stronger transfer learning performance on novel tasks.

Diabetes Mellitus, one of the leading causes of death worldwide, has no cure to date and can lead to severe health complications, such as retinopathy, limb amputation, cardiovascular diseases, and neuronal disease, if left untreated. Consequently, it becomes crucial to take precautionary measures to avoid/predict the occurrence of diabetes. Machine learning approaches have been proposed and evaluated in the literature for diabetes prediction. This paper proposes an IoT-edge-Artificial Intelligence (AI)-blockchain system for diabetes prediction based on risk factors. The proposed system is underpinned by the blockchain to obtain a cohesive view of the risk factors data from patients across different hospitals and to ensure security and privacy of the user's data. Furthermore, we provide a comparative analysis of different medical sensors, devices, and methods to measure and collect the risk factors values in the system. Numerical experiments and comparative analysis were carried out between our proposed system, using the most accurate random forest (RF) model, and the two most used state-of-the-art machine learning approaches, Logistic Regression (LR) and Support Vector Machine (SVM), using three real-life diabetes datasets. The results show that the proposed system using RF predicts diabetes with 4.57% more accuracy on average compared to LR and SVM, with 2.87 times more execution time. Data balancing without feature selection does not show significant improvement. The performance is improved by 1.14% and 0.02% after feature selection for PIMA Indian and Sylhet datasets respectively, while it reduces by 0.89% for MIMIC III.

In this paper, we tackle a novel federated learning (FL) problem for optimizing a family of X-risks, to which no existing FL algorithms are applicable. In particular, the objective has the form of $\mathbb E_{z\sim S_1} f(\mathbb E_{z'\sim S_2} \ell(w; z, z'))$, where two sets of data $S_1, S_2$ are distributed over multiple machines, $\ell(\cdot)$ is a pairwise loss that only depends on the prediction outputs of the input data pairs $(z, z')$, and $f(\cdot)$ is possibly a non-linear non-convex function. This problem has important applications in machine learning, e.g., AUROC maximization with a pairwise loss, and partial AUROC maximization with a compositional loss. The challenges for designing an FL algorithm for X-risks lie in the non-decomposability of the objective over multiple machines and the interdependency between different machines. To this end, we propose an active-passive decomposition framework that decouples the gradient's components with two types, namely active parts and passive parts, where the active parts depend on local data that are computed with the local model and the passive parts depend on other machines that are communicated/computed based on historical models and samples. Under this framework, we develop two provable FL algorithms (FeDXL) for handling linear and nonlinear $f$, respectively, based on federated averaging and merging. We develop a novel theoretical analysis to combat the latency of the passive parts and the interdependency between the local model parameters and the involved data for computing local gradient estimators. We establish both iteration and communication complexities and show that using the historical samples and models for computing the passive parts do not degrade the complexities. We conduct empirical studies of FeDXL for deep AUROC and partial AUROC maximization, and demonstrate their performance compared with several baselines.

While data-driven predictive models are a strictly technological construct, they may operate within a social context in which benign engineering choices entail implicit, indirect and unexpected real-life consequences. Fairness of such systems -- pertaining both to individuals and groups -- is one relevant consideration in this space; it arises when data capture protected characteristics upon which people may be discriminated. To date, this notion has predominantly been studied for a fixed model, often under different classification thresholds, striving to identify and eradicate undesirable, discriminative and possibly unlawful aspects of its operation. Here, we backtrack on this fixed model assumption to propose and explore a novel definition of cross-model fairness where individuals can be harmed when one predictor is chosen ad hoc from a group of equally-well performing models, i.e., in view of utility-based model multiplicity. Since a person may be classified differently across models that are otherwise considered equivalent, this individual could argue for a predictor granting them the most favourable outcome, employing which may have adverse effects on others. We introduce this scenario with a two-dimensional example and linear classification; then, we present a comprehensive empirical study based on real-life predictive models and data sets that are popular with the algorithmic fairness community; finally, we investigate analytical properties of cross-model fairness and its ramifications in a broader context. Our findings suggest that such unfairness can be readily found in the real life and it may be difficult to mitigate by technical means alone as doing so is likely to degrade predictive performance.