Publications

Linear object detection in document images using multiple object tracking

Abstract

Linear objects convey substantial information about document structure, but are challenging to detect accurately because of degradation (curved, erased) or decoration (doubled, dashed). Many approaches can recover some vector representation, but only one closed-source technique introduced in 1994, based on Kalman filters (a particular case of Multiple Object Tracking algorithm), can perform a pixel-accurate instance segmentation of linear objects and enable to selectively remove them from the original image. We aim at re-popularizing this approach and propose: 1. a framework for accurate instance segmentation of linear objects in document images using Multiple Object Tracking (MOT); 2. document image datasets and metrics which enable both vector- and pixel-based evaluation of linear object detection; 3. performance measures of MOT approaches against modern segment detectors; 4. performance measures of various tracking strategies, exhibiting alternatives to the original Kalman filters approach; and 5. an open-source implementation of a detector which can discriminate instances of curved, erased, dashed, intersecting and/or overlapping linear objects.

Continue reading

Metrics for community dynamics applied to unsupervised attacks detection

Abstract

Attack detection in big networks has become a necessity. Yet, with the ever changing threat landscape and massive amount of data to handle, network intrusion detection systems (NIDS) end up being obsolete. Different machine-learning-based solutions have been developed to answer the detection problem for data with evolving statistical distributions. However, no approach has proved to be both scalable and robust to passing time. In this paper, we propose a scalable and unsupervised approach to detect behavioral patterns without prior knowledge on the nature of attacks. For this purpose, we define novel metrics for graph community dynamics and use them as feature with unsupervised detection algorithm on the UGR’16 dataset. The proposed approach improves existing detection algorithms by 285.56% in precision and 222.82% in recall when compared to usual feature extraction (FE) using isolation forest.

Continue reading

Software supply-chain security: Issues and countermeasures

Abstract

Software application development is a complex activity which involves various actors and organizations in what is called the software supply chain. The evolution of the software supply chain led to numerous benefits such as profit maximization, code mutualization, and the optimization of lead times. However, the complexity of the software supply chain results in multiple security issues and attacks because compromises are highly prevalent. An attacker that compromises a single link (e.g., by maliciously modifying the software) in the software supply chain, can harm users of this software and this attack technique is frequently being exploited to attack high profile companies. We can provide a holistic and effective security solution to the software supply chain only if its security state and features are well understood. We discuss how we can achieve strong resilience of the software supply chain to cyberthreats. Next, we propose a holistic end-to-end security approach for the software supply chain.

Continue reading

Learning sentinel-2 reflectance dynamics for data-driven assimilation and forecasting

Abstract

Over the last few years, massive amounts of satellite multispectral and hyperspectral images covering the Earth’s surface have been made publicly available for scientific purpose, for example through the European Copernicus project. Simultaneously, the development of self-supervised learning (SSL) methods has sparked great interest in the remote sensing community, enabling to learn latent representations from unlabeled data to help treating downstream tasks for which there is few annotated examples, such as interpolation, forecasting or unmixing. Following this line, we train a deep learning model inspired from the Koopman operator theory to model long-term reflectance dynamics in an unsupervised way. We show that this trained model, being differentiable, can be used as a prior for data assimilation in a straightforward way. Our datasets, which are composed of Sentinel-2 multispectral image time series, are publicly released with several levels of treatment.

Continue reading

Languages of higher-dimensional timed automata

Abstract

We present a new language semantics for real-time concurrency. Its operational models are higher-dimensional timed automata (HDTAs), a generalization of both higher-dimensional automata and timed automata. We define languages of HDTAs as sets of interval-timed pomsets with interfaces. As an application, we show that language inclusion of HDTAs is undecidable. On the other hand, using a region construction we can show that untimings of HDTA languages have enough regularity so that untimed language inclusion is decidable.

Continue reading

GenIDA, a participatory patient registry for genetic forms of intellectual disability provides detailed caregiver reported information on 237 individuals with Koolen-de Vries syndrome

Abstract

Continue reading

Forecasting electricity prices: An optimize then predict-based approach

Abstract

We are interested in electricity price forecasting at the European scale. The electricity market is ruled by price regulation mechanisms that make it possible to adjust production to demand, as electricity is difficult to store. These mechanisms ensure the highest price for producers, the lowest price for consumers and a zero energy balance by setting day-ahead prices, i.e. prices for the next 24h. Most studies have focused on learning increasingly sophisticated models to predict the next day’s 24 hourly prices for a given zone. However, the zones are interdependent and this last point has hitherto been largely underestimated. In the following, we show that estimating the energy cross-border transfer by solving an optimization problem and integrating it as input of a model improves the performance of the price forecasting for several zones together.

Continue reading

An experience report on the optimization of the product configuration system of Renault

Abstract

The problem of configuring a variability model is widespread in many different domains. A leading automobile manufacturer has developed its technology internally to model vehicle diversity. This technology relies on the approach known as knowledge compilation to explore the configurations space. However, the growing variability and complexity of the vehicles’ range hardens the space representation problem and impacts performance requirements. This paper tackles these issues by exploiting symmetries that represent isomorphic parts in the configurations space. A new method describes how these symmetries are exploited and integrated. The extensive experiments we conducted on datasets from the automobile manufacturer show our approach’s robustness and effectiveness: the achieved gain is a reduction of 52.13% in space representation and 49.81% in processing time on average

Continue reading

Optimization of the product configuration system of Renault

Abstract

The problem of configuring a variability model is widespread in many different domains. Renault has developed its technology internally to model vehicle diversity. This technology relies on the approach known as knowledge compilation to explore the configurations space. However, the growing variability and complexity of the vehicles’ range hardens the space representation problem and impacts performance requirements. This paper tackles these issues by exploiting symmetries that represent isomorphic parts in the configurations space. A new method describes how these symmetries are exploited and integrated. The extensive experiments we conducted on datasets from the automobile manufacturer show our approach’s robustness and effectiveness: the achieved gain is a reduction of 52.13% in space representation on average.

Continue reading

A MOP-based implementation for method combinations

Abstract

In traditional object-oriented languages, the dynamic dispatch algorithm is hardwired to select and execute the most specific method in a polymorphic call. In CLOS, the Common Lisp Object System, an abstraction known as “method combinations” allows the programmer to define their own dispatch scheme. When Common Lisp was standardized, method combinations were not mature enough to be fully specified.In 2018, using SBCL as a research vehicle, we analyzed the unfortunate consequences of this under-specification and proposed a layer on top of method combinations designed to both correct a number of observed behavioral inconsistencies, and propose an extension called “alternative combinators”. Following this work, SBCL underwent a number of internal changes that fixed the reported inconsistencies, although in a way that hindered further experimentation.In this paper, we analyze SBCL’s new method combinations implementation and we propose an alternative design. Our solution is standard-compliant so any Lisp implementation can potentially use it. It is also based on the MOP, meaning that it is extensible, which restores the opportunity for further experimentation. In particular, we revisit our former “alternative combinators” extension, broken after 2018, and demonstrate that provided with this new infrastructure, it can be re-implemented in a much simpler and non-intrusive way.

Continue reading