| 1. |
- Amundin, Mats, et al.
(author)
-
A proposal to use distributional models to analyse dolphin vocalisation
- 2017
-
In: Proceedings of the 1st International Workshop on Vocal Interactivity in-and-between Humans, Animals and Robots, VIHAR 2017. - 9782956202905 ; , s. 31-32
-
Conference paper (peer-reviewed)abstract
- This paper gives a brief introduction to the starting points of an experimental project to study dolphin communicative behaviour using distributional semantics, with methods implemented for the large scale study of human language.
|
|
| 2. |
- Abbas, Nadeem, 1980-, et al.
(author)
-
Smart Forest Observatories Network : A MAPE-K Architecture Based Approach for Detecting and Monitoring Forest Damage
- 2023
-
In: Proceedings of the Conference Digital solutions for detecting and monitoring forest damage.
-
Conference paper (other academic/artistic)abstract
- Forests are essential for life, providing various ecological, social, and economic benefits worldwide. However, one of the main challenges faced by the world is the forest damage caused by biotic and abiotic factors. In any case, the forest damages threaten the environment, biodiversity, and ecosystem. Climate change and anthropogenic activities, such as illegal logging and industrial waste, are among the principal elements contributing to forest damage. To achieve the United Nations' Sustainable Development Goals (SDGs) related to forests and climate change, detecting and analyzing forest damages, and taking appropriate measures to prevent or reduce the damages are essential. To that end, we envision establishing a Smart Forest Observatories (SFOs) network, as shown below, which can be either a local area or a wide area network involving remote forests. The basic idea is to use Monitor, Analyze, Plan, Execute, and Knowledge (MAPE-K) architecture from autonomic computing and self-adaptive software systems domain to design and develop the SFOs network. The SFOs are planned to collect, analyze, and share the collected data and analysis results using state-of-the-art methods. The principal objective of the SFOs network is to provide accurate and real-time data to policymakers and forest managers, enabling them to develop effective policies and management strategies for global forest conservation that help to achieve SDGs related to forests and climate change.
|
|
| 3. |
- Kokkinakis, Dimitrios, 1965
(author)
-
Shallow Features for Differentiating Disease-Treatment Relations using Supervised Learning, a pilot study
- 2009
-
In: Proceedings of the 12th International Conference TSD (Text, Speech and Dialogue). Springer Verlag, LNCS/LNAI series.. ; 5729, s. 395-402
-
Conference paper (other academic/artistic)abstract
- Clinical narratives provide an information rich, nearly unexplored corpus of evidential knowledge that is considered as a challenge for practitioners in the language technology field, particularly because of the nature of the texts (excessive use of terminology, abbreviations, orthographic term variation), the significant opportunities for clinical research that such material can provide and the potentially broad impact that clinical findings may have in every day life. It is therefore recognized that the capability to automatically extract key concepts and their relationships from such data will allow systems to properly understand the content and knowledge embedded in the free text which can be of great value for applications such as information extraction and question & answering. This paper gives a brief presentation of such textual data and its semantic annotation, and discuss the set of semantic relations that can be observed between diseases and treatments in the sample. The problem is then designed as a machine learning task in which the relations are tried to be learned in a supervised fashion, using pre-annotated data. The challenges designing the problem and empirical results are presented.
|
|
| 4. |
- Banda, Francesco, et al.
(author)
-
BIOMASS L2 Prototype Processor : Current Status
- 2019
-
In: International Geoscience and Remote Sensing Symposium (IGARSS). - 9781538691540 ; , s. 5996-5999
-
Conference paper (peer-reviewed)abstract
- The ESA BIOMASS mission will be the 7th Earth Explorer measuring the above-ground biomass (AGB) in the world's forests. The current ESA Level-2 (L2) implementation study focuses on defining and implementing the main algorithms for forest parameter retrieval from BIOMASS data. After the first year of L2 study innovative results were achieved: the development of ground cancellation, in particular, has proved to be huge value, since it removes from the data the effects of environmental variability and contributions unrelated to the forest carried in the ground scattering. In this paper the current processor implementation and validation activities of the L2 team will be described.
|
|
| 5. |
- Forsman, Mona, et al.
(author)
-
Estimation of tree stem attributes using terrestrial photogrammetry
- 2012
-
In: International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. - : Copernicus Gesellschaft. ; 39-B5, s. 261-265
-
Conference paper (other academic/artistic)abstract
- The objective of this work was to create a method to measure stem attributes of standing trees on field plots in the forest using terrestrial photogrammetry. The primary attributes of interest are the position and the diameter at breast height (DBH).The developed method creates point clouds from image from three or more calibrated cameras attached to a calibrated rig. SIFT features in multiple images in combination with epipolar line filtering are used to make high quality matching in images with large amounts of similar features and many occlusions. After projection of the point cloud to a simulated ground plane, RANSAC filtering is applied, followed by circle fitting to the remaining points.To evaluate the algorithm, a camera rig of five Canon digital system cameras with a baseline of 123 cm and up to 40 cm offset in height was constructed. The rig was used in a field campaign at the Remningstorp forest test area in southern Sweden. Ground truth was collected manually by surveying and manual measurements.Initial results show estimated tree stem diameters within 10% of the ground truth. This suggest that terrestrial photogrammetry is a viable method to measure tree stem diameters on circular field plots.
|
|
| 6. |
- Thuvander, Liane, 1970, et al.
(author)
-
Procedural digital twin generation for co-creating in VR focusing on vegetation
- 2022
-
In: International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives. - 1682-1750. ; XLVIII-4/W5-2022, s. 189-196
-
Conference paper (peer-reviewed)abstract
- An early-stage development of a Digital Twin (DT) in Virtual Reality (VR) is presented, aiming for civic engagement in a new urban development located in an area that is a forest today. The area is presently used for recreation. For the developer, it is important both to communicate how the new development will affect the forest and allow for feedback from the citizen. High quality DT models are time-consuming to generate, especially for VR. Current model generation methods require the model developer to manually design the virtual environment. Furthermore, they are not scalable when multiple scenarios are required as a project progresses. This study aimed to create an automated, procedural workflow to generate DT models and visualize large-scale data in VR with a focus on existing green structures as a basis for participatory approaches. Two versions of the VR prototype were developed in close cooperation with the urban developer and evaluated in two user tests. A procedural workflow was developed for generating DT models and integrated into the VR application. For the green structures, efforts focused on the vegetation, such as realistic representation and placement of different types of trees and bushes. Only navigation functions were enabled in the first user test with practitioners (9 participants). Interactive functions were enabled in the second user test with pupils (age 15, 9 participants). In both tests, the researchers observed the participants and carried out short reflective interviews. The user test evaluation focussed on the perception of the vegetation, general perception of the VR environment, interaction, and navigation. The results show that the workflow is effective, and the users appreciate green structure representations in VR environments in both user tests. Based on the workflow, similar scenes can be created for any location in Sweden. Future development needs to concentrate on the refinement of buildings and information content. A challenge will be balancing the level of detail for communication with residents.
|
|
| 7. |
|
|
| 8. |
- Rashid, Maheen, et al.
(author)
-
Equine Pain Behavior Classification via Self-Supervised Disentangled Pose Representation
- 2022
-
In: 2022 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2022). - : Institute of Electrical and Electronics Engineers (IEEE). ; , s. 152-162
-
Conference paper (peer-reviewed)abstract
- Timely detection of horse pain is important for equine welfare. Horses express pain through their facial and body behavior, but may hide signs of pain from unfamiliar human observers. In addition, collecting visual data with detailed annotation of horse behavior and pain state is both cumbersome and not scalable. Consequently, a pragmatic equine pain classification system would use video of the un-observed horse and weak labels. This paper proposes such a method for equine pain classification by using multi-view surveillance video footage of unobserved horses with induced orthopaedic pain, with temporally sparse video level pain labels. To ensure that pain is learned from horse body language alone, we first train a self-supervised generative model to disentangle horse pose from its appearance and background before using the disentangled horse pose latent representation for pain classification. To make best use of the pain labels, we develop a novel loss that formulates pain classification as a multi-instance learning problem. Our method achieves pain classification accuracy better than human expert performance with 60% accuracy. The learned latent horse pose representation is shown to be viewpoint covariant, and disentangled from horse appearance. Qualitative analysis of pain classified segments shows correspondence between the pain symptoms identified by our model, and equine pain scales used in veterinary practice.
|
|
| 9. |
- Ahmed, Faisal, et al.
(author)
-
Machine Learning-Based Tomato Leaf Disease Diagnosis Using Radiomics Features
- 2023
-
In: Proceedings of the Fourth International Conference on Trends in Computational and Cognitive Engineering - TCCE 2022. - : Springer Science and Business Media Deutschland GmbH. - 9789811994821 - 9789811994838 ; , s. 25-35
-
Conference paper (peer-reviewed)abstract
- Tomato leaves can be infected with various infectious viruses and fungal diseases that drastically reduce tomato production and incur a great economic loss. Therefore, tomato leaf disease detection and identification are crucial for maintaining the global demand for tomatoes for a large population. This paper proposes a machine learning-based technique to identify diseases on tomato leaves and classify them into three diseases (Septoria, Yellow Curl Leaf, and Late Blight) and one healthy class. The proposed method extracts radiomics-based features from tomato leaf images and identifies the disease with a gradient boosting classifier. The dataset used in this study consists of 4000 tomato leaf disease images collected from the Plant Village dataset. The experimental results demonstrate the effectiveness and applicability of our proposed method for tomato leaf disease detection and classification.
|
|
| 10. |
- Alshihabi, Omran, et al.
(author)
-
CropSAT – A decision support system for practical use of satellite images in precision agriculture
- 2020
-
In: Lecture Notes in Electrical Engineering. - Cham : Springer International Publishing. - 1876-1100. ; 684, s. 415-421
-
Conference paper (peer-reviewed)abstract
- CropSAT is an interactive decision support system (DSS) that provides vegetation index (VI) maps from Sentinel-2 data all across the globe and lets users delineate fields, design variable-rate application of user specified inputs (mainly nitrogen, but also e.g. fungicides or growth regulators) based on the VI maps. The CropSAT DSS was initially developed in a research project at the Swedish University of Agricultural Sciences (SLU), and has since its launch in 2015 been continuously developed in a private-public-partnership between SLU, private companies and the Swedish Board of Agriculture. Now it has global coverage, is continuously updated with new satellite images, and is provided free-of-charge in multiple languages (including Arabic and French). The present study aims at describing the CropSAT systems, summarizing research results from the ongoing developmental process and pointing to opportunities for applications in precision agriculture, e.g. in Morocco and other countries in North Africa.
|
|
