The topic of this master's thesis is devoted to design and realization of autonomous camera system of the selected object with possibility of remote access. The content of the theoretical part is description of components of the camera system and close analysis of the used software, including the selection of suitable components. The practical part is approaching to the initial installation of the system and verification of the functionality of individual components based on simple commands. The next step is developing an mobile application comunicattting with the camera system based on low data flow.

The monitoring of animal control devices (animal traps) in remote areas currently requires field workers to visit each device on a regular basis, which is costly and time consuming. Better monitoring practices could allow DOC to increase their trapping practices through reduced costs. Essentially, the aim of this paper is to reduce the number of man-hours, and hence resources, required to check each trap. An attempt will be made to use wireless communications to check the status of each trap, and hence decide whether or not it will need to be checked, bringing benefits of efficiency and cost savings to the Department of Conservation. It is recognised that the environment is very difficult for traditional wireless communications to operate reliably and therefore new methods or technologies were investigated for this application. A system operating at 27MHz using a modified pulse position modulation scheme was found to be an appropriate solution; however the success of wireless communications in pest control management is dependent upon the trapping location, patterns and terrain.

In specialized object detection tasks and domains, it is sometimes only possible to collect and annotate a small amount of data for training and evaluation, which constrains training to a low data regime that can lead to poor generalization. In this thesis, the impact of annotations from additional classes, referred to as complementary labels, when learning a target class is studied as a potential approach to improve performance in the low data regime, for object detection. In particular, the thesis aims to investigate in which data regimes complementary labels seem beneficial, whether labels from different complementary classes contribute equally to the performance on the target class, and how varying the number of complementary classes can affect the performance on the target class. Two datasets were studied; CSAW-S, a medical dataset, and MSCOCO, a natural dataset. For each of these datasets, three experiments were conducted to examine various aspects of complementary labels. First, an experiment that compares the use of all available complementary labels and no complementary labels is conducted for various data regimes. Second, an experiment that leaves out individual complementary classes during training is performed. Third, an experiment that varies the number of complementary classes used during training is performed. The results suggest that complementary labels are helpful in the low data regime, provided the complementary classes have sufficient representation in the dataset. Furthermore, complementary classes that have clear context and interaction with the target class seem to be beneficial, and the impact of individual complementary classes does not seem to be cumulative. Lastly, increasing the number of complementary classes used seems to have a stabilizing effect on the target class performance, provided enough classes are used. Due to limitations in the methodology and choice of experiments, these findings are not conclusive. Nevertheless, various improvements to the methodology of studying complementary labels have been identified, which can help future studies present stronger conclusions. / I specialiserade domäner och uppgifter inom objektdetektion är det ibland inte möjligt att samla mer än en liten mängd data för träning och evaluering. Detta kan leda till dålig generalisering av objektdetektorer när ny data påträffas. I detta examensarbete undersöks komplementära etiketter från tillagda klasser som ett potentiellt sätt att förbättra generaliseringen av objektdetektion av en huvudklass. Mer specifikt fokuserar arbetet på att förstå i vilka datamängdsstorlekar som tillagda klasser kan vara användbara för inlärning av huvudklassen, huruvida olika tillagda klasser har lika inflytande på huvudklassen samt hur tillagda klasser påverkar objektdetektorns prestation på huvudklassen när antalet klasser varieras. Två datamängder studerades; CSAW-S, som är en medicinsk datamängd, och MSCOCO, som är en naturlig datamängd. På båda datamängderna genomförs tre experiment som undersöker olika aspekter av tillagda klasser. I det första experimentet jämförs träning av en huvudklass med och utan tillagda klasser med olika mängder träningsdata. I det andra experimentet lämnas individuella tillagda klasser ur träningen. I det tredje experimentet varieras antalet tillagda klasser som används i träningen. Av resultaten föreslås att tillagda klasser är användbara för att öka prestationen på osedd data när träningen begränsas till små datamängder och de tillagda klasserna har tillräcklig representation in datamängden. Utöver detta så verkar de mest fördelaktiga tillagda klasserna vara de som bidrar med tydligt sammanhang och interagerar tydligt med huvudklassen; fördelarna och nackdelarna som enstaka tillagda klasser bidrar med verkar dock inte vara kumulativa. Slutligen verkar prestationen på huvudklassen stabiliseras när antalet tillagda klasser ökar. På grund av begränsningar i metoden och valet av experimenten bör undersökningsresultaten tas som indikationer och inte definita slutsatser. Flera förbättringspunkter har dock identifierats och föreslagits i metoden angående studerandet av tillagda klasser, vilket kan möjliggöra starkare slutsatser i framtida studier.

The Master’s Thesis deals with a ZigBee technology and its devices working each other in wireless personal area network. The ZigBee and its advantages are compared with other wireless protocols working in ISM bands. A first part deals with a topology of IEEE 802.4.15 WPAN and the ZigBee features. Types and format of the ZigBee data-frame are mentioned. A Second part of this thesis describes a design and testing of the ZigBee devices. Results of this thesis are demonstrated on ZMU (ZigBee Modules Utility) program that has been developed for the testing of this technology.

Ces dernières années, le domaine de l’apprentissage profond a connu des progrès énormes dans des applications allant de la génération d’images, détection d’objets, modélisation du langage à la réponse aux questions visuelles. Les approches classiques telles que l’apprentissage supervisé nécessitent de grandes quantités de données étiquetées et spécifiques à la tâches. Cependant, celles-ci sont parfois coûteuses, peu pratiques, ou trop longues à collecter. La modélisation efficace en données, qui comprend des techniques comme l’apprentissage few-shot (à partir de peu d’exemples) et l’apprentissage self-supervised (auto-supervisé), tentent de remédier au manque de données spécifiques à la tâche en exploitant de grandes quantités de données plus “générales”. Les progrès de l’apprentissage profond, et en particulier de l’apprentissage few-shot, s’appuient sur les benchmarks (suites d’évaluation), les métriques d’évaluation et les jeux de données, car ceux-ci sont utilisés pour tester et départager différentes méthodes sur des tâches précises, et identifier l’état de l’art. Cependant, du fait qu’il s’agit de versions idéalisées de la tâche à résoudre, les benchmarks sont rarement équivalents à la tâche originelle, et peuvent avoir plusieurs limitations qui entravent leur rôle de sélection des directions de recherche les plus prometteuses. De plus, la définition de métriques d’évaluation pertinentes peut être difficile, en particulier dans le cas de sorties structurées et en haute dimension, telles que des images, de l’audio, de la parole ou encore du texte. Cette thèse discute des limites et des perspectives des benchmarks existants, des fonctions de coût (training losses) et des métriques d’évaluation (evaluation metrics), en mettant l’accent sur la modélisation générative - les Réseaux Antagonistes Génératifs (GANs) en particulier - et la modélisation efficace des données, qui comprend l’apprentissage few-shot et self-supervised. La première contribution est une discussion de la tâche de modélisation générative, suivie d’une exploration des propriétés théoriques et empiriques des fonctions de coût des GANs. La deuxième contribution est une discussion sur la limitation des few-shot classification benchmarks, certains ne nécessitant pas de généralisation à de nouvelles sémantiques de classe pour être résolus, et la proposition d’une méthode de base pour les résoudre sans étiquettes en phase de testing. La troisième contribution est une revue sur les méthodes few-shot et self-supervised de détection d’objets , qui souligne les limites et directions de recherche prometteuses. Enfin, la quatrième contribution est une méthode efficace en données pour la description de vidéo qui exploite des jeux de données texte et vidéo non supervisés. / In recent years, the field of deep learning has seen tremendous progress for applications ranging from image generation, object detection, language modeling, to visual question answering. Classic approaches such as supervised learning require large amounts of task-specific and labeled data, which may be too expensive, time-consuming, or impractical to collect. Data-efficient methods, such as few-shot and self-supervised learning, attempt to deal with the limited availability of task-specific data by leveraging large amounts of general data. Progress in deep learning, and in particular, few-shot learning, is largely driven by the relevant benchmarks, evaluation metrics, and datasets. They are used to test and compare different methods on a given task, and determine the state-of-the-art. However, due to being idealized versions of the task to solve, benchmarks are rarely equivalent to the original task, and can have several limitations which hinder their role of identifying the most promising research directions. Moreover, defining meaningful evaluation metrics can be challenging, especially in the case of high-dimensional and structured outputs, such as images, audio, speech, or text. This thesis discusses the limitations and perspectives of existing benchmarks, training losses, and evaluation metrics, with a focus on generative modeling—Generative Adversarial Networks (GANs) in particular—and data-efficient modeling, which includes few-shot and self-supervised learning. The first contribution is a discussion of the generative modeling task, followed by an exploration of theoretical and empirical properties of the GAN loss. The second contribution is a discussion of a limitation of few-shot classification benchmarks, which is that they may not require class semantic generalization to be solved, and the proposal of a baseline method for solving them without test-time labels. The third contribution is a survey of few-shot and self-supervised object detection, which points out the limitations and promising future research for the field. Finally, the fourth contribution is a data-efficient method for video captioning, which leverages unsupervised text and video datasets, and explores several multimodal pretraining strategies.