Learning space | EPFL Graph Search

Learning space or learning setting refers to a physical setting for a learning environment, a place in which teaching and learning occur. The term is commonly used as a more definitive alternative to "classroom," but it may also refer to an indoor or outdoor location, either actual or virtual. Learning spaces are highly diverse in use, configuration, location, and educational institution. They support a variety of pedagogies, including quiet study, passive or active learning, kinesthetic or physical learning, vocational learning, experiential learning, and others. As the design of a learning space impacts the learning process, it is deemed important to design a learning space with the learning process in mind. History The word school derives from Greek σχολή (scholē), originally meaning "leisure" and also "that in which leisure is employed", and later "a group to whom lectures were given, school". The Japanese word for school, gakuen, means "learning garden" or "garde

Sustainable Architecture for Educational Space in Africa and China

Ning Liu

Countries in the Global South are producing the most important educational needs of the 21st century. Fast-growing cities can provide the most striking examples of these current tendencies. Just over a decade after the start of the new millennium, the needs for quality-education and appropriate learning space are huge. Yet they are still largely underestimated in many countries of the Global South. With high-level demographical growth and fast-expanding urban boundaries, the developing metropolis provides a fascinating place to analyze societal mutations. School â as one of the fundamental institutions in modern society â deserves timely observation and careful analysis. The research work is dedicated to the study of educational architecture in two African cities Ouagadougou (Burkina Faso) and Cape Town (South Africa) and one Chinese city: Hangzhou. Because of the fundamental relationship between school and society, learning places need to make statement of their urban positions. The hypothesis proposes to question the concept of sustainability in the framework of cultural-environmental qualities between school and city, between users and space. Conducted by an architect, this thesis is the synthesis of a research process together with the professional practice. It is benchmarked by a path of three steps: critical analysis of literature, in situ case-study and cross-cultural comparison with proposal of an assessment method that opens up for future perspectives. The thesis argues that the current time is crucial for our society to reconsider the place of educational architecture in the city. A truly sustainable architecture is transformable in the time-span of its usage and transformative in the changing urban environment. With a discursive comparative view between built cases located in these three cities, this work attempts to provide new insights on the design of the learning space in the urban future. The synthesis provides a panel of cultural-environmental quality profiles as guidelines to future design innovation opportunities.

EPFL2015

Ambient Awareness for the Orchestration of Collaborative Problem Solving

Hamed Seiied Alavi

Classroom orchestration refers to the teacher's (or tutor's) responsibility to identify and manage the evolving learning opportunities and constraints, in real-time. We study the use of computer technology to support this orchestration task in a multi-constraint pedagogical practice: "recitation section". In university teaching, recitation sections complement lecture sessions by providing an opportunity for students to practice the course material in teams, under the supervision of the tutors. We start by analyzing recitation sections as they are held in our university and identify the causes of their inefficiencies. We show that lack of information on the progress and the current status of the students makes the orchestration task difficult for the tutors, and on the students' side results in a hesitation to interact with their peers. Thus, we propose an ambient awareness tool that makes this information available. Our tool, called Lantern, is an interactive lamp which is placed on each group's table that shows its status, namely the specific exercise that the group is currently working on, if assistance is needed, and, if so, for how long this assistance has been needed. Several questions are raised about how Lantern can facilitate the orchestration task and increase collaboration within and between teams. To answer these questions, we describe two user studies. In the first study, we show how Lantern can help the tutors share their resources between teams, enhance students' productivity, increase collaboration among students, and improve the structure of inter-team communication. In the second study, we compare Lantern with another tool, Shelf, which provides the same information but on a single screen in front of the classroom. It appears that Shelf, instead of encouraging collaboration, creates a competitive atmosphere among students. We relate the differences between Lantern and Shelf as to how they position the displayed information in the classroom space. In a second version of Lantern, the lamps are equipped with a wireless communication module that connects them to a server. The server receives every user action and controls the lamps by specifying what should be displayed on each of them. This allows for flexibility both in terms of the variety of the learning scripts that Lantern can implement and also in terms of integration with other technologies inside and outside of the classroom. Finally, we explore the possibilities of using Lantern in other learning contexts, such as high school courses, training workshops, or open learning spaces.

EPFL2012

Human-Centered Scene Understanding via Crowd Counting

Weizhe Liu

Human-centered scene understanding is the process of perceiving and analysing a dynamic scene observed through a network of sensors with emphasis on human-related activities. It includes the visual perception of human-related activities from either single image or video sequence. Scene understanding with focus of human-related activities is becoming increasingly popular which results in the demand of algorithms that can efficiently model crowd activity in different real-world scenarios. In this thesis, we exploit human-centered scene understanding through crowd counting. Counting people is a challenging task due to perspective distortion and occlusion. We tackle these problems by developing algorithms to leverage a variety of data modalities including single image, video sequence and scene perspective map. First, we introduce an end-to-end trainable deep architecture for crowd counting that combines features obtained using multiple receptive field sizes and learns the importance of each such feature at each image location. In other words, our approach adaptively encodes the scale of the contextual information required to accurately predict crowd density. This yields an algorithm that outperforms previous crowd counting methods, especially when perspective effects are strong.Second, we explicitly model the scale changes and reason in terms of people per square-meter. We show that feeding the perspective model to the network allows us to enforce global scale consistency and that this model can be obtained on the fly from the drone sensors. In addition, it also enables us to enforce physically-inspired temporal consistency constraints that do not have to be learned. This yields an algorithm that outperforms previous methods in inferring crowd density from a moving drone camera especially when perspective effects are strong.Third, for video sequence, we advocate estimating people flows across image locations between consecutive images and inferring the people densities from these flows instead of directly regressing them. This enables us to impose much stronger constraints encoding the conservation of the number of people. As a result, it significantly boosts performance without requiring a more complex architecture. Furthermore, it allows us to exploit the correlation between people flow and optical flow to further improve the results. We also show that leveraging people conservation constraints in both a spatial and temporal manner makes it possible to train a deep crowd counting model in an active learning setting with much fewer annotations. This significantly reduces the annotation cost while still leading to similar performance to the full supervision case.

EPFL2021