École Polytechnique Fédérale de Lausanne
Infoscience - École polytechnique fédérale de LausanneNot a member yet
191401 research outputs found
Sort by
Intraday solar irradiance forecasting using public cameras
No full textWith the significant increase in photovoltaic (PV) electricity generation, more attention has been given to PV power forecasting. Indeed, accurate forecasting allows power grid operators to better schedule and dispatch their assets, such as energy storage systems and reserve. In this paper, a hybrid deep learning model and a convolutional neural network with memory is proposed, to provide intraday (2 h) solar irradiance forecasts using sequentially -collected images from public webcams. The performance of the proposed model is compared to those of a standard time -series forecast models, a linear regression as well as state-of-the-art neural networks. All models are trained and tested using images collected from two webcams on EPFL's campus for just over a year. The results show that the proposed method outperforms all other models and matches the state-of-the-art methodology while providing simplicity of implementation and efficient computation.LAPDLODES
Class Specific Feature Disentanglement and Text Embeddings for Multi-label Generalized Zero Shot CXR Classification
No full textRobustness of medical image classification models is limited by its exposure to the candidate disease classes. Generalized zero shot learning (GZSL) aims at correctly predicting seen and unseen classes and most current GZSL approaches have focused on the single label case. It is common for chest x-rays to be labelled with multiple disease classes. We propose a novel multi-label GZSL approach using: 1) class specific feature disentanglement and 2) semantic relationship between disease labels distilled from BERT models pre-trained on biomedical literature. We learn a dictionary from distilled text embeddings, and leverage them to synthesize feature vectors that are representative of multi-label samples. Compared to existing methods, our approach does not require class attribute vectors, which are an essential part of GZSL methods for natural images but are not available for medical images. Our approach outperforms state of the art GZSL methods for chest xray images.LTS
Self-supervised learning-based cervical cytology for the triage of HPV-positive women in resource-limited settings and low-data regime
No full textScreening Papanicolaou test samples has proven to be highly effective in reducing cervical cancer-related mortality. However, the lack of trained cytopathologists hinders its widespread implementation in low-resource settings. Deep learning-assisted telecytology diagnosis emerges as an appealing alternative, but it requires the collection of large annotated training datasets, which is costly and time-consuming. In this paper, we demonstrate that the abundance of unlabeled images that can be extracted from Pap smear test whole slide images presents a fertile ground for self-supervised learning methods, yielding performance improvements compared to off-the-shelf pre-trained models for various downstream tasks. In particular, we propose Cervical Cell Copy-Pasting (C3P) as an effective augmentation method, which enables knowledge transfer from public and labeled single-cell datasets to unlabeled tiles. Not only does C3P outperforms naive transfer from single-cell images, but we also demonstrate its advantageous integration into multiple instance learning methods. Importantly, all our experiments are conducted on our introduced in-house dataset comprising liquid-based cytology Pap smear images obtained using low-cost technologies. This aligns with our long-term objective of deep learning-assisted telecytology for diagnosis in low-resource settings
Geospatial Tools and Remote Sensing Strategies for Timely Humanitarian Response: A Case Study on Drought Monitoring in Eswatini
No full textThis article explores the escalating impact of natural disasters, particularly droughts, in the Southern African Development Community (SADC), with a specific focus on Eswatini. Over the last century, approximately 63 million people in SADC countries have been affected by droughts, leading to challenges in agriculture, livestock losses, and severe food and water shortages. Despite being the smallest SADC nation, the Kingdom of Eswatini faces disproportionate consequences due to its susceptibility to climate variability, particularly drought. The inadequacy and unreliability of rainfall have resulted in a drastic reduction in food production, with maize, a staple crop, experiencing a 70% decline. This adverse trend, spanning three decades, has heightened the vulnerability of farmers to climatic shocks, hindering sustainable agricultural development and impeding poverty alleviation efforts. To address the growing threat of drought in the kingdom, a comprehensive approach is imperative, involving coordinated plans and the development of swift humanitarian relief strategies. This study utilized remote sensing technologies to monitor drought and assess its repercussions, evaluating the impact on agricultural production. Additionally, geospatial tools, including Open Route Service (ORS) and Near Neighbor Analysis algorithms, were employed to optimize humanitarian supply chain logistics. Results from the analysis, including Vegetation Health Index (VHI) fluctuations and drought severity mapping, reveal that 1990 was the year the kingdom was most severely hit by drought. This study also found that smallholder farmers practicing rainfed agriculture in vulnerable regions, such as the lower Middleveld and western Lowveld, suffered the severe socioeconomic consequences of agricultural drought, including income loss, food insecurity, and migration patterns. Through this integrated approach, decision makers can engage in targeted interventions, focusing on farming areas needing irrigation infrastructures or populated areas requiring a coordinated humanitarian response amidst climate variability.SHS-EN
Entre Communication & Participation. Usage du transmedia storytelling en aménagement et urbanisme
No full textColloque final du projet de recherche The narrative making of the city, financé par le Fonds national de la recherche scientifique dans le cadre de l’action COST 18126 —Writing Urban Places. New Narratives of the European CityLAB-
Personalised and Adjustable Interval Type-2 Fuzzy-Based PPG Quality Assessment for the Edge
No full textMost of today's wearable technology provides seamless cardiac activity monitoring. Specifically, the vast majority employ Photoplethysmography (PPG) sensors to acquire blood volume pulse information, which is further analysed to extract useful and physiologically related features. Nevertheless, PPG-based signal reliability presents different challenges that strongly affect such data processing. This is mainly related to the fact of PPG morphological wave distortion due to motion artefacts, which can lead to erroneous interpretation of the extracted cardiac-related features. On this basis, in this paper, we propose a novel personalised and adjustable Interval Type-2 Fuzzy Logic System (IT2FLS) for assessing the quality of PPG signals. The proposed system employs a personalised approach to adapt the IT2FLS parameters to the unique characteristics of each individual's PPG signals. Additionally, the system provides adjustable levels of personalisation, allowing healthcare providers to adjust the system to meet specific requirements for different applications. The proposed system obtained up to 93.72% for average accuracy during validation. The presented system has the potential to enable ultra-low complexity and real-time PPG quality assessment, improving the accuracy and reliability of PPG-based health monitoring systems at the edge.ES
Write What YouWant: Applying Text-to-Video Retrieval to Audiovisual Archives
No full textAudiovisual (AV) archives, as an essential reservoir of our cultural assets, are suffering from the issue of accessibility. The complex nature of the medium itself made processing and interaction an open challenge still in the field of computer vision, multimodal learning, and human-computer interaction, as well as in culture and heritage. In recent years, with the raising of video retrieval tasks, methods in retrieving video content with natural language (text-to-video retrieval) gained quite some attention and have reached a performance level where real-world application is on the horizon. Appealing as it may sound, such methods focus on retrieving videos using plain visual-focused descriptions of what has happened in the video and finding videos such as instructions. It is too early to say such methods would be the new paradigms for accessing and encoding complex video content into high-dimensional data, but they are indeed innovative attempts and foundations to build future exploratory interfaces for AV archives (e.g., allow users to write stories and retrieve related snippets in the archive, or encoding video content at high-level for visualisation). This work filled the application gap by examining such text-tovideo retrieval methods from an implementation point of view and proposed and verified a classifier-enhanced workflow to allow better results when dealing with in-situ queries that might have been different from the training dataset. Such a workflow is then applied to the real-world archive from Television Suisse Romande (RTS) to create a demo. At last, a humancentred evaluation is conducted to understand whether the text-to-video retrieval methods improve the overall experience of accessing AV archives.EMPLU
Using Video Streaming Feeds to Encourage Informal Learning
No full textSocial media have become an indispensable part of daily life, particularly among university students, who regularly browse social news feeds in their spare time. Due to their pervasiveness, social media platforms provide an opportunity for influencing user behavior and encouraging informal learning. In this paper, we present an experiment using an online video recommendation application designed to blend micro-informative content with general content according to user preferences and activity history. Based on a one-week study, we conclude that injecting micro-informative content into video streaming platforms has the potential to improve the perceived satisfaction of users and can act as a potential catalyst to motivate users to consume more informative content online.SCI-STI-D
