Offer DescriptionThis position entails working in the project idealcv-cm, a collaborative project funded by the local government of Madrid, Spain.Despite significant advances in deep learning-based computer vision systems, many limitations still exist.
The main objective of this project is to advance the state of the art of deep learning-based computer vision systems, improving their accuracy, robustness, efficiency, and providing better explainability.
Our hypothesis is that "the whole is more than the sum of its parts," such that achieving significant advances beyond marginal progress requires working simultaneously on multiple fronts within the deep learning and visual data representation paradigms.In particular, we propose to take a holistic approach, addressing objectives at three distinct levels, with the intention of contributing to the entire system, and we anticipate that improvements at each level will combine synergistically to produce significant advances.
At the low level, we plan to target the core of deep neural networks (DNNs) and propose a new bio-inspired artificial neuron model, improving the building blocks of deep learning systems.
At the mid-level, we will focus on improving the visual representation of data (whether input, internal, or output) by designing new deep architectures and processing layers.
Finally, at the high level, we will focus on designing new learning methods to train deep neural networks with imperfect visual data, exploring innovative methods to learn from data in the wild, in scenarios with unannotated data, where annotations are noisy or not aligned with the task at hand (e.g., different granularity), or in new domains with little data.Tasks to be performed:Translating biophysical properties of biological dendrites into new artificial neuron models.Integrating these models into artificial neural networks.Validation of these new networks as models of neuronal activity for the retina and brain.Location: Instituto de Óptica, Madrid, Spain.Salary: Monthly gross of 1797.20€Requirements:Master's degree in Physics, Engineering, or Computer Science.Experience in optics and vision.Experience in artificial neural networks.Excellent level of English.Application Process:Send application letter, CV, and contact information of 3 academic references to: ****** Fields:PhysicsEngineeringCommunication SciencesEducation Level: Master Degree or equivalentLanguages: English - ExcellentWork Location:Instituto de Óptica "Daza de Valdés", Madrid, Spain.Number of offers available: 1Contact:E-Mail: ****** #J-18808-Ljbffr