Scientific direction Development of key enabling technologies
Transfer of knowledge to industry

PhD : selection by topics

See all positions [+]

Learning a class parameterized detector with 3D models

Département Intelligence Ambiante et Systèmes Interactifs (LIST)

Laboratoire Vision et Apprentissage pour l'analyse de scènes

01-10-2021

SL-DRT-21-1071

bertrand.luvison@cea.fr

Artificial intelligence & Data intelligence (.pdf)

Object detectors have been signicantly improved thanks to deep neural network. It is now quite simple to detect on object as far as images with this kind of object in various conditions are available. These datasets exist for some classes (21 classes in PascalVOC, 80 in COCO, etc) but in practive, it is usually necessary to create a specific dataset for classes of interest and then to finetune or adapt an already learnt network. The constitution of this dataset is extremely tedious, time-consuming. Moreover, adding to detect a new class for such model is not flexible at all since you need to fully learn from scratch with the extended dataset. The challenge of this PhD is to free the detector from this constraint of fix set of detectable classes. Such a formulation could be seen as "Zero-shot learning" but this kind of modelling relies on an auxiliary semantic knowledge such as attribut definition. In this way any image can be described through these attributes and semantical definition closer to the one found in dictionnaries can be made event on class of object never seen. The objective of this PhD is not to tackle this problem through "Zero-shot learning" perspective but rather to provide the list of object type, one wants to detect using 3D models. The aim is not build a representative semantic ontology of the world but rather to learn a neural network to find in an image, whichever defined 3D model. In the way old "template matching" methods worked, the aim is now to use the strength of neural network to process this task efficiently on several classes of object simultaneously.

Download the offer (.zip)

Unsupervised monocular moving object detection using spatio-temporal constraints

Département Intelligence Ambiante et Systèmes Interactifs (LIST)

Laboratoire Vision et Apprentissage pour l'analyse de scènes

01-09-2021

SL-DRT-21-1072

camille.dupont@cea.fr

Artificial intelligence & Data intelligence (.pdf)

Traditional object detection techniques based on supervised learning require a large amount of training data. This observation has motivated the scientific community to investigate learning methods with the lowest possible supervision. In this context, there is a growing popularity for self-supervised learning which automatically generates training annotations by exploiting the relationships between input signals. Among the principal tasks that can benefit from learning, we note especially low-level tasks with little semantics such as optical flow estimation, depth estimation from monocular video or pre-learning tasks. The objective of this thesis is to go even further in the exploitation of this learning family in an object detection context. In particular, the work expected in the thesis is to design an object detection model trained from monocular video data which can then be used in the image or video to detect all the object classes present in training data.

Download the offer (.zip)

Towards a better understanding of the degradation mechanisms for solid oxide cells: a multiscale modeling approach coupling ab-initio computations and elementary kinetic simulations

Département Thermique Conversion et Hydrogène (LITEN)

Laboratoire essais et systèmes

01-11-2021

SL-DRT-21-1078

maxime.hubert@cea.fr

Advanced hydrogen and fuel-cells solutions for energy transition (.pdf)

Solid oxide cells (SOCs) are electrochemical devices operating at high temperature that can directly convert fuel into electricity (fuel cell mode ? SOFC) or electricity into fuel (electrolysis mode ? SOEC). In recent years, the interest on SOCs has grown significantly thanks to their wide range of technological applications that could offer innovative solutions for the transition toward a renewable energy market. However, despite of all their advantages, the SOCs lifetime is still insufficient to envisage the industrial deployment of this technology. Indeed, the SOCs durability remains limited by various degradation phenomena, which mainly occur at both electrodes. Nevertheless, the underlying mechanisms and the driving force for the degradation are still not precisely known and they depend on complex mutiphysic phenomena involving different scales. In the frame of the thesis, a mutiphysic and multiscale modeling approach will be proposed for a better understating of the electrodes operation and degradation. For this purpose, ab-initio computations will be developed and then coupled to elementary kinetic models. The simulations will allow unravelling the entangled phenomena involved in the degradation and proposed innovative solutions to enhance the SOCs lifetime.

Download the offer (.zip)

Deep learning for privacy-oriented user profiling

Département Intelligence Ambiante et Systèmes Interactifs (LIST)

Laboratoire Analyse Sémantique Textes et Images

01-12-2021

SL-DRT-21-1126

adrian.popescu@cea.fr

Artificial intelligence & Data intelligence (.pdf)

A wide majority of online platforms implement a business model in which access to services is free in exchange for the right to exploit user data. User profiles are created from shared data and exploited to personalize the proposed services (search results, recommendations, etc.) and thus improve their effectiveness. Two important downsides of the current business model are that (1) its economic performance depends on the level of detail of the data and (2) profiles are created by the platforms and stored on the server side. This leads to important privacy and transparency problems since users have little control over their shared data and the profiles derived from them. Moreover, intrusiveness will increase with the widespread usage of AI, techniques to infer actionable information from raw user data. Alternative solutions are necessary in order to take into account users' expectations in terms of personal data protection and to comply with the European regulatory framework (GDPR1, Digital Services Act2, Artificial Intelligence Act3). The main objective of this PhD is to propose a privacy- and transparency-oriented user profiling methodology, while maintaining the advantages of online service personalization. The fulfillment of this objective entails the creation and storage of user profiles on their devices. These profiles will never be exposed to the services which exploit them because the personalisation step will also be realized at the edge. The mobilization of recent advancements in several areas of incremental learning is necessary : 1. Semi-supervised learning ? in order to learn effective deep models using a low amount of manually annotated data which are available for the application domains. 2. Transfer learning ? in order to leverage the strength of deep representations learned with large datasets, while also adapting these representations to the applicative context. 3. On-device learning ? in order to infer knowledge uniquely on the users' mobile devices. 4. Multimedia learning ? in order to integrate data of different types (images, vidéos, texts, geolocation etc.) The PhD will build on existing works done by CEA LIST LASTI and will explore axes such as: 1. Exploiting the models proposed by domain experts in order to optimize the performance of deep models learned in a semi-supervised manner. 2. Adapting the local inferences for each user by combining public and personal data for training whenever such a combination is useful. 3. Proposing deep model compression methods which are optimized for different data types and a variety of mobile devices. 4. Providing the users with effective profile understanding control mechanisms in order to increase their trust in the profiling and, ultimately, personalization processes.

Download the offer (.zip)

Impact of InGaN quantum wells design and composition on the performance of micro-LEDs

Département d'Optronique (LETI)

Laboratoire des Composants Emissifs

01-11-2020

SL-DRT-21-1131

fabian.rol@cea.fr

Emerging materials and processes for nanotechnologies and microelectronics (.pdf)

The growth and fabrication of InGaN/GaN blue LEDs have reached a high degree of maturity thanks to their extensive use in the field of lighting. Thanks to their robustness and high efficiency, these LEDs are seen as promising candidates to make high-luminance and high resolution micro-displays for the field of augmented reality. However, the size of µLEDs composing the pixels of micro-displays are of a few micrometers. At these dimensions, the non-radiative recombination of electrons and holes occurring on the defect present at the sidewalls becomes dominant and degrade the performances. The passivation of these defects is the usual way to recover a higher efficiency. In addition, the lateral diffusion of carriers in InGaN quantum wells (QW) should also play a role on the performance of µLEDs by controlling the migration of electron and holes to the defective sidewalls. Over the last 10 years, researchers have gained a much better understanding of the physics of InGaN QW in large LEDs and they have optimized the epi-structures accordingly. However, going from macro to micro-LEDs have created new constraints (mainly the non-radiative recombination at the sidewalls). For this PhD, we propose to study the impact of InGaN QW design and composition on carrier diffusion properties and their impact on the performance of µLEDs. The student will be in charge of growing quantum wells and complete LEDs structures by MOCVD and will characterize them thoroughly. A large part of the PhD will be dedicated to the spectroscopic and electro-optical characterization of the samples that will be completed by structural characterization.

Download the offer (.zip)

137 Results found (Page 23 of 23)
first   previous  19 - 20 - 21 - 22 - 23

See all positions