Postdoctoral position or Research Engineer at ENIB on multimodal learning for underwater object detection and classification

Job title:

Postdoctoral position or Research Engineer at ENIB on multimodal learning for underwater object detection and classification

Company

Job description

Offer DescriptionDuration: 12 months.Desired hiring date: the position is open from October 15, 2024. The call will remain open until satisfactory candidate is found.Take-home salary: 2100€/month.Workplace: Lab-STICC UMR CNRS, Brest National School of Engineering, Brittany, France.Context of the postdoctoral position:The Lab-STICC is a research unit of the French national center for scientific research (CNRS). The staffs (more than 650 people) are located over the different institutes on several geographical sites in Brittany working within one central theme: “from sensor to knowledge”.We are seeking an outstanding postdoctoral research fellow with experience in computer vision and machine learning to work on a project investigating the design of algorithms of deep multimodal learning for underwater object detection and classification. The postdoctoral position is funded under the research project ASTRID ROV-Chasseur supported by the French Research Funding Agency (ANR). The ROV-Chasseur project gathers the research units of Lab-STICC (Brest), LIS (Toulon) and LIFO (Orleans) around the development of an intelligent ROV (Remotely Operated underwater Vehicle) to efficiently detect and classify specific underwater objects as fish and mines. In particular, the project is interested to technological challenges in the development and deployment of deep models on the platform of an ROV working in an underwater environment.Objectives and challenges:Automatic underwater object recognition can be divided into two steps: (1) specific object detection, which aims to localize every single object in the underwater image, (2) and object classification, which aims to identify the class or category of each detected object. Object recognition in underwater images is an open challenge in pattern recognition due to the difficulties presented by the underwater environment. For example, the luminosity changes frequently because of the ocean current, the visibility is limited and the complex background sometimes changes rapidly due to moving aquatic plants. When dealing with fish species recognition, the fish move freely in all directions, and they can also hide behind rocks and algae. In addition, the problems of fish overlapping and of the similarity in shape and patterns among fish of different species pose significant challenges in this application. Within this project, we are interested in studying multimodal deep learning techniques for integrating multiple sources of information to significantly improve the performance of underwater object recognition systems. To deal with the lack of a large amount of annotated underwater data for model training, we will investigate weakly supervised learning techniques. The postdoctoral research fellow shall focus on the design of efficient deep models that consider the combination of heterogeneous visual cues from two modalities: sound and optical images/videos. This involves addressing the real-world multimodal learning challenges, namely: multimodal data representation, multimodal fusion, and multimodal alignment. To handle the multiple classes issue, a hierarchical classification could be performed to classify objects first into family levels and then into species categories. This work will advance understanding of deep neural networks used in underwater object detection and classification and will bring more clarity to the decision-making mechanisms of these systems.

Sub-topics/subtasks include:– A state of the art and an understanding of the deep learning methods proposed for multimodal object recognition.– Designing robust multimodal deep learning architectures for underwater object recognition (detection and classification) by combining two types of data: RGB and sound images/videos.– Implementation of a validation protocol with performances evaluation (in terms of size of the training set, training time, processing time, precision and recall…). The functional prototype could be shared with the other partners of the project to provide first in situ results.– Delivery of results on fish and mines recognition.– Comparison performances (drawing strengths and weaknesses) between the different proposed frameworks.– Publishing of the results and participation in dissemination activities.– Redaction of technical reports and code documentation.– Participation in project monitoring.Required qualifications and skills:Ph.D. degree in computer science or related field with specialization in computer vision/machine learning.Experience within deep learning frameworks is highly recommended.Knowledge of programming languages: C++, Python, MATLAB.Knowledge of libraries: PyTorch, TensorFlow, OpenCV, PCL, Open3D.Proficient in English language (written and oral).Interpersonal skills and the ability to work in a multidisciplinary team are recommended.Taste for research activities in submarine applications.To apply:If your interests are compatible, please feel free to send the following information in a single PDF file to :– Detailed curriculum vitae, including list of publications.– Cover letter explaining your interest and how your experience fits with this postdoctoral position.– One-page research statement.– Names and contact details of at least two referees.Where to apply E-mailbenzinou@enib.frRequirementsResearch Field Computer science » Systems design Education Level PhD or equivalentAdditional InformationWork Location(s)Number of offers available 1 Company/Institute ENIB Country France City Brest Postal Code 29200 GeofieldContact CityPLOUZANE WebsiteStreet945 avenue du Technopôle E-Mailbenzinou@enib.frSTATUS: EXPIREDShare this page

Expected salary

Location

Brest, Finistère

Job date

Sat, 05 Oct 2024 07:02:54 GMT

To help us track our recruitment effort, please indicate in your email/cover letter where (hiring-jobs.com) you saw this job posting.