Active learning | EPFL Graph Search

Active learning is "a method of learning in which students are actively or experientially involved in the learning process and where there are different levels of active learning, depending on student involvement." states that "students participate [in active learning] when they are doing something besides passively listening." According to Hanson and Moser (2003) using active teaching techniques in the classroom can create better academic outcomes for students. Scheyvens, Griffin, Jocoy, Liu, & Bradford (2008) further noted that “by utilizing learning strategies that can include small-group work, role-play and simulations, data collection and analysis, active learning is purported to increase student interest and motivation and to build students ‘critical thinking, problem-solving and social skills”. In a report from the Association for the Study of Higher Education, authors discuss a variety of methodologies for promoting active learning. They cite literature that indicates studen

Learning strategies and representations for intuitive robot learning from demonstration

Thibaut Antoine Kulak

Robots are becoming more and more present around us, both in industries and in our homes. One key capability of robots is their adaptability to various situations that might appear in the real world. Robot skill learning is therefore a crucial aspect of robotics aiming to provide robots with programs enabling them to perform one or several tasks successfully. While such programming is usually done by an engineer or a developer, making robot programming available to anyone would dramatically increase the range of applications currently feasible for robots. Learning from Demonstration (LfD) is a robot skill learning paradigm addressing this aim by developing intuitive frameworks for non-expert users to easily (re)program robots. While Learning from Demonstration has emerged as a successful way to program robots, several limitations remain to be addressed. Typical approaches still require some forms of preprocessing, such as the alignment of the demonstrations, or the choice of the movement representation. Also, the algorithms have to run with a relatively low number of demonstrations that human users are typically willing to give, while being performant, adaptable and generalizable to new situations. In this thesis, we propose to address these shortcomings with methods that make Learning from Demonstration more intuitive and user-friendly. We notably propose a novel movement representation requiring no demonstration alignment, and active learning strategies that permit to learn complex skills from fewer demonstrations.

EPFL2021

Level up the interaction space with Cellulo robots

Lucas Burget

Cellulo, as tabletop tangible robots, are a great tool for interactive learning and upper limb rehabilita-tion. For the moment, the main way of interaction is to take them in hand and move them in the plane.Thus, due to the laws of physics, the different gestures induced are mainly constrained to be in 2D andthe workspace is bounded by the length of the arm. More precisely, for upper limb rehabilitation, itlimits the solicitation of certain arm joints, for instance the shoulder absduction/adduction. In orderto overcome these limitations, the idea was to use two wristbands with sensors to consider the armsof the user as a remote control. For the rehabilitation part, this would allow the patient to train thejoints of his arm with greater amplitude. But more generally, this new mean of interaction with theCellulo allows more possibility in the creation of activities.This semester project studied two different aspects: pointing direction and gesture recognition. Thefirst could be used by the game Master to change the behavior of some robots in order to adapt to thesituation. And the second could be destined to control a swarm of robots. The governing principle ofthis semester project was to implement systems as light and easy to use as possible.

2019

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

Human-Centered Scene Understanding via Crowd Counting

Weizhe Liu

EPFL2021