This thesis explores motion trajectory-based approaches to recognize human actions in real-world, unconstrained videos. Recognizing actions is an important task in applications such as video retrieval, surveillance, human-robot interactions, analysis of sports videos, summarization of videos, behaviour monitoring, etc. There has been a considerable amount of research done in this regard. Earlier work used to be on videos captured by static cameras where it was relatively easy to recognise the actions. With more videos being captured by moving cameras, recognition of actions in such videos with irregular camera motion is still a challenge in unconstrained settings with variations in scale, view, illumination, occlusion and unrelated motions in the background. With the increase in videos being captured from wearable or head-mounted cameras, recognizing actions in egocentric videos is also explored in this thesis. At first, an effective motion segmentation method to identify the camera motion in videos captured by moving cameras is explored. Next, action recognition in videos captured in normal third-person view (perspective) is discussed. Further, the action recognition approaches for first-person (egocentric) views are investigated. First-person videos are often associated with frequent unintended camera motion. This is due to the motion of the head resulting in the motion of the head-mounted cameras (wearable cameras). This is followed by recognition of actions in egocentric videos in a multicamera setting. And lastly, novel feature encoding and subvolume sampling (for “deep” approaches) techniques are explored in the context of action recognition in videos. The first part of the thesis explores two effective segmentation approaches to identify the motion due to camera. The first approach is based on curve fitting of the motion trajectories and finding the model which best fits the camera motion model. The curve fitting approach works when the trajectories generated are smooth enough. To overcome this drawback and segment trajectories under non-smooth conditions, a second approach based on trajectory scoring and grouping is proposed. By identifying the instantaneous dominant background motion and accordingly aggregating the scores (denoting the “foregroundness”) along the trajectory, the motion that is associated with the camera can be separated from the motion due to foreground objects. Additionally, the segmentation result has been used to align videos from moving cameras, resulting in videos that seem to be captured by nearly-static cameras. In the second part of the thesis, recognising actions in normal videos captured from third-person cameras is investigated. To this end, two kinds of descriptors are explored. The first descriptor is the covariance descriptor adapted for the motion trajectories. The covariance descriptor for a trajectory encodes the co-variations of different features along the trajectory’s length. Covariance, being a second-order encoding, encodes information of the trajectory that is different from that of the first-order encoding. The second descriptor is based on Granger causality. The novel causality descriptor encodes the “cause and effect” relationships between the motion trajectories of the actions. This type of interaction descriptors captures the causal inter-dependencies among the motion trajectories and encodes complimentary information different from those descriptors based on the occurrence of features. The causal dependencies are traditionally computed on time-varying signals. We extend it further to capture dependencies between spatiotemporal signals and compute generalised causality descriptors which perform better than their traditional counterparts. An egocentric or first-person video is captured from the perspective of the personof-interest (POI). The POI wears a camera and moves around doing his/her activities. This camera records the events and activities as seen by him/her. The POI who is performing actions or activities is not seen by the camera worn by him/her. Activities performed by the POI are called first-person actions and third-person actions are those done by others and observed by the POI. The third part of the thesis explores action recognition in egocentric videos. Differentiating first-person and third-person actions is important when summarising/analysing the behaviour of the POI. Thus, the goal is to recognise the action and the perspective from which it is being observed. Trajectory descriptors are adapted to recognise actions along with the motion trajectory ranking method of segmentation as pre-processing step to identify the camera motion. The motion segmentation step is necessary to remove unintended head motion (camera motion) during video capture. To recognise actions and corresponding perspectives in a multi-camera setup, a novel inter-view causality descriptor based on the causal dependencies between trajectories in different views is explored. Since this is a new problem being addressed, two first-person datasets are created with eight actions in third-person and first-person perspectives. The first dataset is a single camera dataset with action instances from first-person and third-person views. The second dataset is a multi-camera dataset with each action instance having multiple first-person and third-person views. In the final part of the thesis, a feature encoding scheme and a subvolume sampling scheme for recognising actions in videos is proposed. The proposed Hyper-Fisher Vector feature encoding is based on embedding the Bag-of-Words encoding into the Fisher Vector encoding. The resulting encoding is simple, effective and improves the classification performance over the state-of-the-art techniques. This encoding can be used in place of the traditional Fisher Vector encoding in other recognition approaches. The proposed subvolume sampling scheme, used to generate second layer features in “deep” approaches for action recognition in videos, is based on iteratively increasing the size of the valid subvolumes in the temporal direction to generate newer subvolumes. The proposed sampling requires lesser number of subvolumes to be generated to “better represent” the actions and thus, is less computationally intensive compared to the original sampling scheme. The techniques are evaluated on large-scale, challenging, publicly available datasets. The Hyper-Fisher Vector combined with the proposed sampling scheme perform better than the state-of-the-art techniques for action classification in videos.

Una red egocéntrica o personal está compuesta por un ego, quien es el actor principal y sus alters, personas con quienes se relaciona. Estos estudios centran su investigación en los vínculos relacionales, de modo que el análisis de la red es óptimo para describir y comprender la influencia de los alters en el desarrollo de conductas relacionadas con la salud Objetivo: Describir las características (composición, estructura y soporte) de las redes egocéntricas de mujeres en edad fértil que acuden al servicio de Ginecología de un hospital nacional de Lima. Materiales y métodos: Se encuestó a 73 mujeres en edad fértil que acudieron al servicio de Ginecología del Hospital Nacional Dos de Mayo durante los meses febrero-marzo del 2020, siendo 71 las analizadas. Resultados: Las redes estuvieron compuestas principalmente por alters con características sociodemográficas similares a las del ego. Las redes fueron de 8 alters en promedio y se encontró una gran conexión entre los alters. Así mismo, se encontró que las redes de los egos están conectadas por un familiar. Se encontró que el 81,7% de las mujeres recurrieron a una sola fuente de información, siendo en su mayoría el centro de salud. Conclusiones: Las redes de mujeres en edad fértil muestran patrones repetitivos que consisten en la interrelación con personas que poseen similares características sociodemográficas. Así mismo, los resultados muestran que tanto las redes familiares como los centros de salud pueden ser determinantes para que las mujeres en edad fértil adopten comportamientos preventivos sobre su salud ginecológica. / An egocentric or personal network is composed of an ego, who is the main actor, and it alters people with whom it relates. These studies focus their research on relational ties, so that network analysis is optimal for describing and understanding the influence of alters on the development of health-related behaviors Objective: To describe the characteristics (composition, structure, and support) of egocentric networks of women of childbearing age attending the gynecology service of a national hospital in Lima. Materials and methods: Seventy-one women of childbearing age who attended the gynecology service of the Hospital Nacional Dos de Mayo during February-March 2020 were surveyed. Results: The networks were mainly composed of alters with sociodemographic characteristics similar to those of the ego. The networks were of 8 alters on average and a strong connection between alters was found. Likewise, the egos' networks were found to be connected by a family member. It was found that 81.7% of the women resorted to only one source of information, being mostly the health center. Conclusions: The networks of women of childbearing age show repetitive patterns consisting of interrelationships with people who have similar sociodemographic characteristics. Likewise, the results show that both family networks and health centers can be determinants for women of childbearing age to adopt preventive behaviors regarding their gynecological health. / Tesis

No description available.

No description available.

<p>Navigation is a very important area of spatial information research that presents researchers with a number of challenges. One of these challenges concerns the nature of spatial information encoding itself: is such encoding the result of a single mechanism system, a two-mechanism system or possibly a mixed system? One possible avenue of insight into this problem centers on the disorientation effect as described in Wang & Spelke (2000). A quick survey of basic findings, terminating with Waller & Hodgson (2006), indicates that there seem to be two systems at work. Moreover, the results obtained are based upon experiments carried out in actual reality. A virtual reality experiment was designed in an attempt to replicate the findings described in Waller & Hodgson (2006). The experiment is described in detail and its results are presented. These were found to be sufficiently reliable to justify pointing to a potentially rich field for future research, including such techniques as combining VR with fMRI to achieve more fine-grained results that cannot currently be obtained from the direct use of actual reality only. Underlying factors such as experimental control and data presentation are briefly described in the discussion section.</p> / Master of Science (MS)

<p>Multiple reference frames are used to interact with our surroundings. When these reference frames are in conflict, processing errors can occur. For tactile stimuli, this conflict is highlighted when the hands are crossed over the midline of the body. In this posture, vibrotactile temporal order judgments (TOJs) presented to the hands are impaired compared to an uncrossed posture. This decrease in temporal processing is known as the crossed-hands deficit. The deficit was explored in depth throughout this thesis. In Chapters 2, 3 and 4 different elements of the crossed-hands deficit were evaluated including its connections to the rod and frame test, individual and sex differences within the TOJ task, as well as the influence of vision and body position. These elements were framed with underlying goal of investigating the root cause of the deficit. The data presented here provided evidence for a conflict model of crossed hands processing. A conflict between the internal and external reference frames produced the deficit in temporal processing when the hands were crossed. The role of the body’s midline in understanding multisensory integration was further considered in Chapter 5 through the rubber hand illusion, which is a visuotactile phenomenon whereby an unseen real hand is mislocalized towards a seen rubber hand. When the real hand, rubber hand, or both were crossed over the midline the illusion did not occur. It was hypothesized that a failure to integrate the tactile information presented to the real hand with the visual rubber hand was responsible for the absence of the illusion. Taken together, the data presented in this thesis contribute to the greater understanding of how reference frame conflicts are resolved, particularly when the conflict occurs across the body’s midline.</p> / Doctor of Philosophy (PhD)

No description available.

No description available.

No description available.

Recent advances in camera technology have made it possible to build a comfortable, wearable system which can capture the scene in front of the user throughout the day. Products based on this technology, such as GoPro and Google Glass, have generated substantial interest. In this thesis, I present my work on egocentric vision, which leverages wearable camera technology and provides a new line of attack on classical computer vision problems such as object categorization and activity recognition. The dominant paradigm for object and activity recognition over the last decade has been based on using the web. In this paradigm, in order to learn a model for an object category like coffee jar, various images of that object type are fetched from the web (e.g. through Google image search), features are extracted and then classifiers are learned. This paradigm has led to great advances in the field and has produced state-of-the-art results for object recognition. However, it has two main shortcomings: a) objects on the web appear in isolation and they miss the context of daily usage; and b) web data does not represent what we see every day. In this thesis, I demonstrate that egocentric vision can address these limitations as an alternative paradigm. I will demonstrate that contextual cues and the actions of a user can be exploited in an egocentric vision system to learn models of objects under very weak supervision. In addition, I will show that measurements of a subject's gaze during object manipulation tasks can provide novel feature representations to support activity recognition. Moving beyond surface-level categorization, I will showcase a method for automatically discovering object state changes during actions, and an approach to building descriptive models of social interactions between groups of individuals. These new capabilities for egocentric video analysis will enable new applications in life logging, elder care, human-robot interaction, developmental screening, augmented reality and social media.