462 research outputs found
An API For Building Artificial Worlds For Machine Learning Using Blender
No full textTrabalho de projeto de mestrado, Informática, 2022, Universidade de Lisboa, Faculdade de CiênciasVirtual technologies are a facet of advancement that permeates almost all major fields in our world,
from physics simulations to videogames and movies, urban design and engineering to rocket science,
all make use of this wonderful and powerful tool. One such field is Machine Learning. This project aims
to leverage said virtual technologies to help such a field in the shape of an object identification model
being developed by Thibault Nicolas Langlois.
This is done by using a tool such as Blender and a language such as Python to create an API that
leverages both in order to facilitate the procedural generation of virtual worlds. By generating said
worlds and taking snapshots of them with the rendering power of Blender, this project aims to help
generate vast amounts of data that can be catalogued and sent to feed said models.
To do this, the API was constructed in a way that allows for modular adaptation to whatever purpose it
is to be used for, all while containing an example focused on city generation. This can help guide others
intending on using the API, on how to adapt it to their use case.
This proved to be extremely complex and arduous, requiring many hours of work and lines of code
written. The scale of the endeavour was vast and required a lot to reach the state it currently presents
itself in. It also presented how capable and versatile the technology can be, giving many a tool to the
user’s disposal, as well as the developer.
This proves the path forward for testing many aspects of products or experiments is indeed virtualization
as it has shown to be capable enough in providing the tools necessary to emulate even the most complex
scenarios, all while potentially saving numerous amounts of investment and time, depending on the scale
of the products taking advantage of it
David Thibault, Sr.
No full textSubject: Formal portrait of Arkansas author David Thibault, Sr. Gift of subject's son, David Thibault, 1994. 1. Authors--Arkansas. I. Thibault, David, donor
Features for the classification and clustering of music in symbolic format
No full textTese de mestrado, Engenharia Informática, Universidade de Lisboa, Faculdade de Ciências, 2008Este documento descreve o trabalho realizado no âmbito da disciplina de Projecto em Engenharia Informática do Mestrado em Engenharia Informática da Faculdade de Ciências da Universidade de Lisboa. Recuperação de Informação Musical é, hoje em dia, um ramo altamente activo de investigação e desenvolvimento na área de ciência da computação, e incide em diversos tópicos, incluindo a classificação musical por géneros. O trabalho apresentado centra-se na Classificação de Pistas e de Géneros de música armazenada usando o formato MIDI. Para resolver o problema da classificação de pistas MIDI, extraimos um conjunto de descritores que são usados para treinar um classificador implementado através de uma técnica de Máquinas de Aprendizagem, Redes Neuronais, com base nas notas, e durações destas, que descrevem cada faixa. As faixas são classificadas em seis categorias: Melody (Melodia), Harmony (Harmonia), Bass (Baixo) e Drums (Bateria). Para caracterizar o conteúdo musical de cada faixa, um vector de descritores numérico, normalmente conhecido como ”shallow structure description”, é extraído. Em seguida, eles são utilizados no classificador — Neural Network — que foi implementado no ambiente Matlab. Na Classificação por Géneros, duas propostas foram usadas: Modelação de Linguagem, na qual uma matriz de transição de probabilidades é criada para cada tipo de pista midi (Melodia, Harmonia, Baixo e Bateria) e também para cada género; e Redes Neuronais, em que um vector de descritores numéricos é extraído de cada pista, e é processado num Classificador baseado numa Rede Neuronal. Seis Colectâneas de Musica no formato Midi, de seis géneros diferentes, Blues, Country, Jazz, Metal, Punk e Rock, foram formadas para efectuar as experiências. Estes géneros foram escolhidos por partilharem os mesmos instrumentos, na sua maioria, como por exemplo, baixo, bateria, piano ou guitarra. Estes géneros também partilham algumas características entre si, para que a classificação não seja trivial, e para que a robustez dos classificadores seja testada. As experiências de Classificação de Pistas Midi, nas quais foram testados, numa primeira abordagem, todos os descritores, e numa segunda abordagem, os melhores descritores, mostrando que o uso de todos os descritores é uma abordagem errada, uma vez que existem descritores que confundem o classificador. Provou-se que a melhor maneira, neste contexto, de se classificar estas faixas MIDI é utilizar descritores cuidadosamente seleccionados. As experiências de Classificação por Géneros, mostraram que os Classificadores por Instrumentos (Single-Instrument) obtiveram os melhores resultados. Quatro géneros, Jazz, Country, Metal e Punk, obtiveram resultados de classificação com sucesso acima dos 80%
O trabalho futuro inclui: algoritmos genéticos para a selecção de melhores descritores; estruturar pistas e musicas; fundir todos os classificadores desenvolvidos num único classificador.This document describes the work carried out under the discipline of Computing Engineering Project of the Computer Engineering Master, Sciences Faculty of the Lisbon University. Music Information Retrieval is, nowadays, a highly active branch of research and development in the computer science field, and focuses several topics, including music genre classification. The work presented in this paper focus on Track and Genre Classification of music stored using MIDI format, To address the problem of MIDI track classification, we extract a set of descriptors that are used to train a classifier implemented by a Neural Network, based on the pitch levels and durations that describe each track. Tracks are classified into four classes: Melody, Harmony, Bass and Drums. In order to characterize the musical content from each track, a vector of numeric descriptors, normally known as shallow structure description, is extracted. Then they are used as inputs for the classifier which was implemented in the Matlab environment. In the Genre Classification task, two approaches are used: Language Modeling, in which a transition probabilities matrix is created for each type of track (Melody, Harmony, Bass and Drums) and also for each genre; and an approach based on Neural Networks, where a vector of numeric descriptors is extracted from each track (Melody, Harmony, Bass and Drums) and fed to a Neural Network Classifier. Six MIDI Music Corpora were assembled for the experiments, from six different genres, Blues, Country, Jazz, Metal, Punk and Rock. These genres were selected because all of them have the same base instruments, such as bass, drums, piano or guitar. Also, the genres chosen share some characteristics between them, so that the classification isn’t trivial, and tests the classifiers robustness. Track Classification experiments using all descriptors and best descriptors were made, showing that using all descriptors is a wrong approach, as there are descriptors which confuse the classifier. Using carefully selected descriptors proved to be the best way to classify these MIDI tracks. Genre Classification experiments showed that the Single-Instrument Classifiers achieved the best results. Four genres achieved higher than 80% success rates: Jazz, Country, Metal and Punk. Future work includes: genetic algorithms; structurize tracks and songs; merge all presented classifiers into one full Automatic Genre Classification System
Vers une évaluation de ressources terminologiques individu-centrées
No full text@conference{cn-beust-2005-2, author = {Beust, Pierre and Roy, Thibault}, title = {Vers une évaluation de ressources terminologiques individu-centrées}, year = {2005}, booktitle = {Conférence Terminologie et Intelligence Artificielle (TIA 2005), Communication dans l'atelier " Evaluation des outils terminologiques "}, address = {Rouen, France}
Une plate-forme logicielle dédiée à la cartographie thématique de corpus
No full text@inproceedings{cn-roy-2005, author = {Roy, Thibault}, title = {Une plate-forme logicielle dédiée à la cartographie thématique de corpus}, year = {2005}, month = {juin}, pages = {545--554}, booktitle = {Actes de la 12e conférence sur le traitement automatique des langues naturelles}, organization = {Dourdan, France, 6 au 10 juin 2005}, note = {ISBN 2-9524255-0-7} }National audienc
La relation au monde dans Les Thibault de Roger Martin du Gard
No full textL ensemble romanesque Les Thibault de Roger Martin du Gard (publié de 1922 à 1940) est sous-tendu par un témoignage que l auteur adresse au lecteur sur le monde réel dans lequel l écrivain français vécut. L objectif de cette thèse est d analyser ce témoignage à travers certains aspects des relations multilatérales entre ces trois pôles que sont la société, l œuvre et l auteur. L œuvre romanesque de Roger Martin du Gard informe la société de son époque autant qu elle est informée par elle, et l écrivain, façonné par la société, façonne à son tour une société romanesque. La relation au monde des personnages des Thibault puise ses sources dans l expérience même de l auteur : relations familiales, relations entre les deux sexes, relations sociales et religieuses, autant de clés pour comprendre l œuvre mais peut-être aussi la vie de l auteur. Pour permettre d apprécier la réception de l œuvre de Roger Martin du Gard auprès du public chinois, la dernière partie de cette thèse analyse les deux traductions des Thibault publiées dans les années 1980 en Chine, et propose aussi quelques éléments de comparaison avec une fresque romanesque chinoise à portée elle aussi familiale et sociale, la trilogie de Ba Jin (Pa Kin) intitulée Torrent (Famille Printemps Automne, 1932-1940). L étude des analogies entre ces deux œuvres certes très différentes permet de découvrir comment le roman de Roger Martin du Gard et son message dépassent les frontières, trouvent écho dans d autres cultures, et finissent par rejoindre l humanité universelle.In the eight novels Les Thibault by Roger Martin du Gard (published between 1922 to 1940), the author addresses the reader and describes the world in which the former lives. The objective of this thesis is to examine Martin du Gard s discourse with respect to some aspects of the multilateral relationships between society, author and oeuvre. R.M.G s novels informed society of his time and were themselves shaped by society. The author moulded by society then constructs a fictional world. The relationships in Les Thibault are autobiographical in nature whether familial, gender, social or religious and these constitute the principal means of understanding the oeuvre and the author s own life.To evaluate the reception of Roger Martin du Gard s work among the Chinese public, the final part of this thesis analyses the two translations of Les Thibault, published in China in the 1980s. Comparison with a Chinese trilogy, Torrent (Family Spring Autumn, 1932-1940), by Ba Jin (Pa Kin), although very different but which also considers family and social relationships, allows consideration of how Roger Martin du Gard s message crosses frontiers and reflects universal concerns, those of humanity itself.ARRAS-Bib.electronique (620419901) / SudocSudocFranceF
L’église de Saint-Thibault-des-Vignes et son décor sculpté
No full textThe church of Saint-Thibault-des-Vignes and its sculpted decoration.
The church of Saint-Thibault is little known, and deserves better recognition in the historical context of the building and decoration of romanesque churches in the Ile-de-France region. Its construction dates back to the pilgrimages which began after the return of the relics of Saint Thibault, a saint born at Provins around 1030, who died in Italy in 1066 and was canonized by Pope Alexander II. The church underwent several modifications between the late eleventh century and the Revolution.
The initial construction was apparently the result of a single building campaign ; this is confirmed by the decoration of the capitals of the columns. Two of them are sculpted with geometrical motifs whilst a second group, in the nave, has embellished storied representations or plant motifs. A third group in the apse has capitals which are exclusively carved with plant motifs. The main interest of these capitals lies in the sculpted representations of biblical scenes. Some of these, for example the resurrection of Lazarus, have been mutilated and require better identification.
There were probably two sculptors, perhaps three.
A comparison with the storied capitals of Sainte-Geneviève in Paris is revealing. Despite the slightness of the affinities, the author of this article believes that the sculptor of the main group of storied capitals at Saint-Thibault-des-Vignes could well be the sculptor of the four capitals of Sainte-Geneviève, held today by the Cluny Museum in Paris and probably dating from the late eleventh or early twelfth century.Didier Marie-Hélène. L’église de Saint-Thibault-des-Vignes et son décor sculpté. In: Histoire de l'art, N°15, 1991. Varia. pp. 33-42
Life and literature in Les Thibault by Roger Martin du Gard
No full textL’ensemble romanesque Les Thibault de Roger Martin du Gard (publié de 1922 à 1940) est sous-tendu par un témoignage que l’auteur adresse au lecteur sur le monde réel dans lequel l’écrivain français vécut. L’objectif de cette thèse est d’analyser ce témoignage à travers certains aspects des relations multilatérales entre ces trois pôles que sont la société, l’œuvre et l’auteur. L’œuvre romanesque de Roger Martin du Gard informe la société de son époque autant qu’elle est informée par elle, et l’écrivain, façonné par la société, façonne à son tour une société romanesque. La relation au monde des personnages des Thibault puise ses sources dans l’expérience même de l’auteur : relations familiales, relations entre les deux sexes, relations sociales et religieuses, autant de clés pour comprendre l’œuvre mais peut-être aussi la vie de l’auteur. Pour permettre d’apprécier la réception de l’œuvre de Roger Martin du Gard auprès du public chinois, la dernière partie de cette thèse analyse les deux traductions des Thibault publiées dans les années 1980 en Chine, et propose aussi quelques éléments de comparaison avec une fresque romanesque chinoise à portée elle aussi familiale et sociale, la trilogie de Ba Jin (Pa Kin) intitulée Torrent (Famille – Printemps – Automne, 1932-1940). L’étude des analogies entre ces deux œuvres certes très différentes permet de découvrir comment le roman de Roger Martin du Gard et son message dépassent les frontières, trouvent écho dans d’autres cultures, et finissent par rejoindre l’humanité universelle.In the eight novels Les Thibault by Roger Martin du Gard (published between 1922 to 1940), the author addresses the reader and describes the world in which the former lives. The objective of this thesis is to examine Martin du Gard’s discourse with respect to some aspects of the multilateral relationships between society, author and oeuvre. R.M.G’s novels informed society of his time and were themselves shaped by society. The author moulded by society then constructs a fictional world. The relationships in Les Thibault are autobiographical in nature whether familial, gender, social or religious and these constitute the principal means of understanding the oeuvre and the author’s own life.To evaluate the reception of Roger Martin du Gard’s work among the Chinese public, the final part of this thesis analyses the two translations of Les Thibault, published in China in the 1980s. Comparison with a Chinese trilogy, Torrent (Family – Spring – Autumn, 1932-1940), by Ba Jin (Pa Kin), although very different but which also considers family and social relationships, allows consideration of how Roger Martin du Gard’s message crosses frontiers and reflects universal concerns, those of humanity itself
User Preferences for Access to Textual Information
No full text@inproceedings{ai-roy-2006, author = {Roy, Thibault and Ferrari, Stéphane}, title = {User Preferences for Access to Textual Information}, booktitle = {Proceedings of the First International Workshop on Semantic Media Adaptation and Personalization (SMAP 2006)}, year = {2006}, month = {décembre}, organization = {4 et 5 décembre 2006}, note = {à paraître}, address = {Athènes, Grèce} }International audienc
A tool for thematic cartography of corpora
No full text@inproceedings{ai-roy-2005, author = {Roy, Thibault and Beust, Pierre}, title = {A tool for thematic cartography of corpora}, booktitle = {Proceedings of the International Workshops on Computer Aided Systems Theory (EUROCAST)}, year = {2005}, month = {février}, organization = {7 au 11 février 2005}, pages = {123--126}, address = {Las Palmas de Gran Canaria, Espagne} }International audienc
- …
