Learning to recognize faces incrementally

[EN]Most face recognition systems are based on some form of batch learning. Online face recognition is not only more practical, it is also much more biologically plausible. Typical batch learners aim at minimizing both training error and (a measure of) hypothesis complexity. We show that the same minimization can be done incrementally as long as some form of ”scaffolding” is applied throughout the learning process. Scaffolding means: make the system learn from samples that are neither too easy nor too difficult at each step. We note that such learning behavior is also biologically plausible. Experiments using large sequences of facial images support the theoretical claims. The proposed method compares well with other, numerical calculus-based online learners.

Padrões de ratreamento na visualização de faces próprias e não próprias

Dissertação (mestrado)—Universidade de Brasília, Instituto de Psicologia, Departamento de Processos Psicológicos Básicos, Programa de Pós-Graduação em Ciências do Comportamento, 2016.

Prototype and proof of concept for a display system in retail stores

Selling devices on retail stores comes with the big challenge of grabbing the customer’s attention. Nowadays people have a lot of offers at their disposal and new marketing techniques must emerge to differentiate the products. When it comes to smartphones and tablets, those devices can make the difference by themselves, if we use their computing power and capabilities to create something unique and interactive. With that in mind, three prototypes were developed during an internship: a face recognition based Customer Detection, a face tracking solution with an Avatar and interactive cross-app Guides. All three revealed to have potential to be differentiating solutions in a retail store, not only raising the chance of a customer taking notice of the device but also of interacting with them to learn more about their features. The results were meant to be only proof of concepts and therefore were not tested in the real world.

Audio-Visual Information Processing across different age groups

People possess different sensory modalities to detect, interpret, and efficiently act upon various events in a complex and dynamic environment (Fetsch, DeAngelis, & Angelaki, 2013). Much empirical work has been done to understand the interplay of modalities (e.g. audio-visual interactions, see Calvert, Spence, & Stein, 2004). On the one hand, integration of multimodal input as a functional principle of the brain enables the versatile and coherent perception of the environment (Lewkowicz & Ghazanfar, 2009). On the other hand, sensory integration does not necessarily mean that input from modalities is always weighted equally (Ernst, 2008). Rather, when two or more modalities are stimulated concurrently, one often finds one modality dominating over another. Study 1 and 2 of the dissertation addressed the developmental trajectory of sensory dominance. In both studies, 6-year-olds, 9-year-olds, and adults were tested in order to examine sensory (audio-visual) dominance across different age groups. In Study 3, sensory dominance was put into an applied context by examining verbal and visual overshadowing effects among 4- to 6-year olds performing a face recognition task. The results of Study 1 and Study 2 support default auditory dominance in young children as proposed by Napolitano and Sloutsky (2004) that persists up to 6 years of age. For 9-year-olds, results on privileged modality processing were inconsistent. Whereas visual dominance was revealed in Study 1, privileged auditory processing was revealed in Study 2. Among adults, a visual dominance was observed in Study 1, which has also been demonstrated in preceding studies (see Spence, Parise, & Chen, 2012). No sensory dominance was revealed in Study 2 for adults. Potential explanations are discussed. Study 3 referred to verbal and visual overshadowing effects in 4- to 6-year-olds. The aim was to examine whether verbalization (i.e., verbally describing a previously seen face), or visualization (i.e., drawing the seen face) might affect later face recognition. No effect of visualization on recognition accuracy was revealed. As opposed to a verbal overshadowing effect, a verbal facilitation effect occurred. Moreover, verbal intelligence was a significant predictor for recognition accuracy in the verbalization group but not in the control group. This suggests that strengthening verbal intelligence in children can pay off in non-verbal domains as well, which might have educational implications.

Molecular probes for monitoring mitochondrial movement and function

This thesis explores two distinct parts of mitochondrial physiology: the role of mitochondria in generation of reactive oxygen species (ROS) and mitochondrial morphology and dynamics within cells. The first area of research is covered in Chapters 1-8. Mitochondrial biofunctionality and ROS production are discussed in Chapter 1, followed by the strategy of targeting bioactive compounds to mitochondria by linking them to lipophilic triphenylphosphonium cations (TPP) (Chapter 2). ROS sensors relevant to the research are reviewed in Chapter 3. Chapter 4 presents design and synthesis of novel probes for superoxide detection in mitochondria (MitoNeo-D), cytosol (Neo-D) and extracellular environment (ExCellNeo-D). The results of biological validation of MitoNeo-D and Neo-D performed in the MRC MBU in Cambridge are presented in Chapter 5. A dicationic hydrogen peroxide sensor that utilizes in situ click chemistry is discussed in Chapter 6. Preliminary work on the synthesis of mitochondria-targeted superoxide generators, which led to the development of mitochondria-targeted analogue of paraquat, MitoPQ, is presented in Chapter 7. A set of bifunctional probes (BCN-Mal, BCN-E-BCN and Mito-iTag) for assessing the redox states of protein thiols is discussed in Chapter 8 along with their biological validation. The second part of the thesis is aimed at the study of mitochondrial morphology and dynamics and is presented in Chapters 9-11. Chapter 9 provides background on the classes of fluorophores relevant to the research, the phenomenon of fluorescence quenching and the principle of photoactivation with examples of photoactivatable fluorophores. Next, the background on mitochondrial morphology and heterogeneity is presented in Chapter 10, followed by the ways of imaging and tracking mitochondria within cells by conventional fluorophores and by photoactivatable fluorophores exploiting super-resolution microscopy. Chapter 11 presents the design and synthesis of four photoactivatable fluorophores for mitochondrial tracking, MitoPhotoRhod110, MitoPhotoNIR, Photo-E+, MitoPhoto-E+, along with results of biological validation of MitoPhotoNIR. The results and discussion concludes with Chapter 12, which is a summary and suggestions for future work, followed by the chemistry experimental procedures (Chapter 13), materials and methods for biological experiments (Chapter 14) and references.

Variational and learning models for image and time series inverse problems

Inverse problems are at the core of many challenging applications. Variational and learning models provide estimated solutions of inverse problems as the outcome of specific reconstruction maps. In the variational approach, the result of the reconstruction map is the solution of a regularized minimization problem encoding information on the acquisition process and prior knowledge on the solution. In the learning approach, the reconstruction map is a parametric function whose parameters are identified by solving a minimization problem depending on a large set of data. In this thesis, we go beyond this apparent dichotomy between variational and learning models and we show they can be harmoniously merged in unified hybrid frameworks preserving their main advantages. We develop several highly efficient methods based on both these model-driven and data-driven strategies, for which we provide a detailed convergence analysis. The arising algorithms are applied to solve inverse problems involving images and time series. For each task, we show the proposed schemes improve the performances of many other existing methods in terms of both computational burden and quality of the solution. In the first part, we focus on gradient-based regularized variational models which are shown to be effective for segmentation purposes and thermal and medical image enhancement. We consider gradient sparsity-promoting regularized models for which we develop different strategies to estimate the regularization strength. Furthermore, we introduce a novel gradient-based Plug-and-Play convergent scheme considering a deep learning based denoiser trained on the gradient domain. In the second part, we address the tasks of natural image deblurring, image and video super resolution microscopy and positioning time series prediction, through deep learning based methods. We boost the performances of supervised, such as trained convolutional and recurrent networks, and unsupervised deep learning strategies, such as Deep Image Prior, by penalizing the losses with handcrafted regularization terms.

Fluorogenic hyaluronan-based probe: characterization and use in advanced microscopy

Among all, the application of nanomaterials in biomedical research and most recently in the environmental one has opened the fields of nanomedicine and nanoremediation. Sensing methods based on fluorescence optical probe are generally requested for their selectivity, sensitivity. However, most imaging methods in literature rely on a fluorescent covalent labelling of the system. Therefore, the main aim of this project was to synthetise a biocompatible fluorogenic hyaluronan probe (HA) polymer functionalised with a rhomadine B (RB) moieties and study its behaviour as an optical probe with different materials with microscopy techniques. A derivatization of HA with RB (HA-RB) was successfully obtained providing a photophysical characterization showing a particular fluorescence mechanism of the probe. Firstly, we tested the interaction with different lab-grade micro and nanoplastics in water. Thanks to the peculiar photophysical behaviour of the probe nanoplastics can be detected with confocal microscopy and more interestingly their nature can be discriminated based on the fluorescence lifetime decay with FLIM microscopy. After, the interaction of a model plant derived metabolic enzyme GAPC1 undergoing oxidative-triggered aggregation was explored with the HA-RB. We highlighted the probe interaction with the protein even at early stage of the kinetic. Moreover, nanoparticle tracking analysis (NTA) experiment demonstrates that the probe is in fact able to interact with the small pre-aggregates in the early stage of the aggregation kinetic. Ultimately, we focused on the possibility to apply the probe in a super resolution microscopy technique, PALM, exploiting its aspecific interaction to characterize the surface topography of PTFE polydisperse microplastics. Optimal conditions were reached at high concentration of the probe (70 nM) where 0.5-5 nM is always advisable for this technique. Thanks to the polymeric nature and fluorescence mechanism of the probe, this technique was able to reveal features of PTFE surface under the diffraction limit (< 250 nm).

Quantitative susceptibility mapping as biomarker of neurodegeneration: methodological models and clinical applications

Quantitative Susceptibility Mapping (QSM) is an advanced magnetic resonance technique that can quantify in vivo biomarkers of pathology, such as alteration in iron and myelin concentration. It allows for the comparison of magnetic susceptibility properties within and between different subject groups. In this thesis, QSM acquisition and processing pipeline are discussed, together with clinical and methodological applications of QSM to neurodegeneration. In designing the studies, significant emphasis was placed on results reproducibility and interpretability. The first project focuses on the investigation of cortical regions in amyotrophic lateral sclerosis. By examining various histogram susceptibility properties, a pattern of increased iron content was revealed in patients with amyotrophic lateral sclerosis compared to controls and other neurodegenerative disorders. Moreover, there was a correlation between susceptibility and upper motor neuron impairment, particularly in patients experiencing rapid disease progression. Similarly, in the second application, QSM was used to examine cortical and sub-cortical areas in individuals with myotonic dystrophy type 1. The thalamus and brainstem were identified as structures of interest, with relevant correlations with clinical and laboratory data such as neurological evaluation and sleep records. In the third project, a robust pipeline for assessing radiomic susceptibility-based features reliability was implemented within a cohort of patients with multiple sclerosis and healthy controls. Lastly, a deep learning super-resolution model was applied to QSM images of healthy controls. The employed model demonstrated excellent generalization abilities and outperformed traditional up-sampling methods, without requiring a customized re-training. Across the three disorders investigated, it was evident that QSM is capable of distinguishing between patient groups and healthy controls while establishing correlations between imaging measurements and clinical data. These studies lay the foundation for future research, with the ultimate goal of achieving earlier and less invasive diagnoses of neurodegenerative disorders within the context of personalized medicine.

Smart Mirror: remote verification of an identity document holder

Nowadays, some activities, such as subscribing an insurance policy or opening a bank account, are possible by navigating through a web page or a downloadable application. Since the user is often “hidden” behind a monitor or a smartphone, it is necessary a solution able to guarantee about their identity. Companies are often requiring the submission of a “proof-of-identity”, which usually consists in a picture of an identity document of the user, together with a picture or a brief video of themselves. This work describes a system whose purpose is the automation of these kinds of verifications.

Transiting exoplanets from the CoRoT space mission VIII. CoRoT-7b: the first super-Earth with measured radius

Aims. We report the discovery of very shallow (Delta F/F approximate to 3.4 x 10(-4)), periodic dips in the light curve of an active V = 11.7 G9V star observed by the CoRoT satellite, which we interpret as caused by a transiting companion. We describe the 3-colour CoRoT data and complementary ground-based observations that support the planetary nature of the companion. Methods. We used CoRoT colours information, good angular resolution ground-based photometric observations in- and out- of transit, adaptive optics imaging, near-infrared spectroscopy, and preliminary results from radial velocity measurements, to test the diluted eclipsing binary scenarios. The parameters of the host star were derived from optical spectra, which were then combined with the CoRoT light curve to derive parameters of the companion. Results. We examined all conceivable cases of false positives carefully, and all the tests support the planetary hypothesis. Blends with separation >0.40 '' or triple systems are almost excluded with a 8 x 10(-4) risk left. We conclude that, inasmuch we have been exhaustive, we have discovered a planetary companion, named CoRoT-7b, for which we derive a period of 0.853 59 +/- 3 x 10(-5) day and a radius of R(p) = 1.68 +/- 0.09 R(Earth). Analysis of preliminary radial velocity data yields an upper limit of 21 M(Earth) for the companion mass, supporting the finding. Conclusions. CoRoT-7b is very likely the first Super-Earth with a measured radius. This object illustrates what will probably become a common situation with missions such as Kepler, namely the need to establish the planetary origin of transits in the absence of a firm radial velocity detection and mass measurement. The composition of CoRoT-7b remains loosely constrained without a precise mass. A very high surface temperature on its irradiated face, approximate to 1800-2600 K at the substellar point, and a very low one, approximate to 50 K, on its dark face assuming no atmosphere, have been derived.

Spectral peak resolution and speech recognition in quiet: Normal hearing, hearing impaired and cochlear implant listeners

Spectral peak resolution was investigated in normal hearing (NH), hearing impaired (HI), and cochlear implant (CI) listeners. The task involved discriminating between two rippled noise stimuli in which the frequency positions of the log-spaced peaks and valleys were interchanged. The ripple spacing was varied adaptively from 0.13 to 11.31 ripples/octave, and the minimum ripple spacing at which a reversal in peak and trough positions could be detected was determined as the spectral peak resolution threshold for each listener. Spectral peak resolution was best, on average, in NH listeners, poorest in CI listeners, and intermediate for HI listeners. There was a significant relationship between spectral peak resolution and both vowel and consonant recognition in quiet across the three listener groups. The results indicate that the degree of spectral peak resolution required for accurate vowel and consonant recognition in quiet backgrounds is around 4 ripples/octave, and that spectral peak resolution poorer than around 1–2 ripples/octave may result in highly degraded speech recognition. These results suggest that efforts to improve spectral peak resolution for HI and CI users may lead to improved speech recognition

Serotonergic modulation of face-emotion recognition

Facial expressions of basic emotions have been widely used to investigate the neural substrates of emotion processing, but little is known about the exact meaning of subjective changes provoked by perceiving facial expressions. Our assumption was that fearful faces would be related to the processing of potential threats, whereas angry faces would be related to the processing of proximal threats. Experimental studies have suggested that serotonin modulates the brain processes underlying defensive responses to environmental threats, facilitating risk assessment behavior elicited by potential threats and inhibiting fight or flight responses to proximal threats. In order to test these predictions about the relationship between fearful and angry faces and defensive behaviors, we carried out a review of the literature about the effects of pharmacological probes that affect 5-HT-mediated neurotransmission on the perception of emotional faces. The hypothesis that angry faces would be processed as a proximal threat and that, as a consequence, their recognition would be impaired by an increase in 5-HT function was not supported by the results reviewed. In contrast, most of the studies that evaluated the behavioral effects of serotonin challenges showed that increased 5-HT neurotransmission facilitates the recognition of fearful faces, whereas its decrease impairs the same performance. These results agree with the hypothesis that fearful faces are processed as potential threats and that 5-HT enhances this brain processing.

Emotion recognition: The role of featural and configural face information

Several studies investigated the role of featural and configural information when processing facial identity. A lot less is known about their contribution to emotion recognition. In this study, we addressed this issue by inducing either a featural or a configural processing strategy (Experiment 1) and by investigating the attentional strategies in response to emotional expressions (Experiment 2). In Experiment 1, participants identified emotional expressions in faces that were presented in three different versions (intact, blurred, and scrambled) and in two orientations (upright and inverted). Blurred faces contain mainly configural information, and scrambled faces contain mainly featural information. Inversion is known to selectively hinder configural processing. Analyses of the discriminability measure (A′) and response times (RTs) revealed that configural processing plays a more prominent role in expression recognition than featural processing, but their relative contribution varies depending on the emotion. In Experiment 2, we qualified these differences between emotions by investigating the relative importance of specific features by means of eye movements. Participants had to match intact expressions with the emotional cues that preceded the stimulus. The analysis of eye movements confirmed that the recognition of different emotions rely on different types of information. While the mouth is important for the detection of happiness and fear, the eyes are more relevant for anger, fear, and sadness.

High-resolution AMS (super 14) C dating of post-bomb peat archives of atmospheric pollutants.

Peat deposits in Greenland and Denmark were investigated to show that high-resolution dating of these archives of atmospheric deposition can be provided for the last 50 years by radiocarbon dating using the atmospheric bomb pulse. (super 14) C was determined in macrofossils from sequential one cm slices using accelerator mass spectrometry (AMS). Values were calibrated with a general-purpose curve derived from annually averaged atmospheric (super 14) CO (sub 2) values in the northernmost northern hemisphere (NNH, 30 degrees -90 degrees N). We present a through review of (super 14) C bomb-pulse data from the NNH including our own measurements made in tree rings and seeds from Arizona as well as other previously published data. We show that our general-purpose calibration curve is valid for the whole NNH producing accurate dates within 1-2 years. In consequence, (super 14) C AMS can precisely date individual points in recent peat deposits within the range of the bomb-pulse (from the mid-1950s on). Comparing the (super 14) C AMS results with the customary dating method for recent peat profiles by (super 210) Pb, we show that the use of (super 137) Cs to validate and correct (super 210) Pb dates proves to be more problematic than previously supposed. As a unique example of our technique, we show how this chronometer can be applied to identify temporal changes in Hg concentrations from Danish and Greenland peat cores.