Shuqing Teresa Yeo

Related publications (5)

3D Common Corruptions and Data Augmentation

Andrei Atanov, Oguzhan Fatih Kar, Amir Roshan Zamir, Shuqing Teresa Yeo

We introduce a set of image transformations that can be used as corruptions to evaluate the robustness of models as well as data augmentation mechanisms for training neural networks. The primary distinction of the proposed transformations is that, unlike existing approaches such as Common Corruptions [27], the geometry of the scene is incorporated in the transformations - thus leading to corruptions that are more likely to occur in the real world. We also introduce a set of semantic corruptions (e.g. natural object occlusions. See Fig. 1). We show these transformations are 'efficient' (can be computed on-the-fly), 'extendable' (can be applied on most image datasets), expose vulnerability of existing models, and can effectively make models more robust when employed as '3D data augmentation' mechanisms. The evaluations on several tasks and datasets suggest incorporating 3D information into benchmarking and training opens up a promising direction for robustness research.

IEEE COMPUTER SOC2022

, ,

We present a method for making neural network predictions robust to shifts from the training data distribution. The proposed method is based on making predictions via a diverse set of cues (called 'middle domains') and ensembling them into one strong prediction. The premise of the idea is that predictions made via different cues respond differently to a distribution shift, hence one should be able to merge them into one robust final prediction. We perform the merging in a straightforward but principled manner based on the uncertainty associated with each prediction. The evaluations are performed using multiple tasks and datasets (Taskonomy, Replica, ImageNet, CIFAR) under a wide range of adversarial and non-adversarial distribution shifts which demonstrate the proposed method is considerably more robust than its standard learning counterpart, conventional deep ensembles, and several other baselines.

IEEE2021

Applying IDC theory to education in the Alps region: A response to Chan et al.’s contribution

Pierre Dillenbourg, Kevin Gonyop Kim, Jauwairia Nasir, Jennifer Kaitlyn Olsen, Shuqing Teresa Yeo

In this paper, we present a response to the Interest-Driven Creator (IDC) Theory from a European perspective. Specifically, we raise six questions intended to start a dialogue with respect to IDC Theory’s placement in existing learning theories, its adoption in educational systems, and how it can be influenced by emerging learning technologies and digitalization, which is currently a driving force in the Alps region. By referring to our own work in vocational education and classroom orchestration, we demonstrate how IDC can begin to play a part in guiding innovations and its potential impact on education both in and outside of Asia. With respect to digitalization, rather than allowing technological innovations to fully guide educational decisions, we call for IDC theory to be part of the conversation to help guide future educational designs.

2019

Iterative Classroom Teaching

Thibault Lucien Christian Asselborn, Volkan Cevher, Pierre Dillenbourg, Louis Pierre Faucon, Kamalaruban Parameswaran, Shuqing Teresa Yeo

We consider the machine teaching problem in a classroom-like setting wherein the teacher has to deliver the same examples to a diverse group of students. Their diversity stems from differences in their initial internal states as well as their learning rates. We prove that a teacher with full knowledge about the learning dynamics of the students can teach a target concept to the entire classroom using O (min{d,N} log 1/eps) examples, where d is the ambient dimension of the problem, N is the number of learners, and eps is the accuracy parameter. We show the robustness of our teaching strategy when the teacher has limited knowledge of the learners' internal dynamics as provided by a noisy oracle. Further, we study the trade-off between the learners' workload and the teacher's cost in teaching the target concept. Our experiments validate our theoretical results and suggest that appropriately partitioning the classroom into homogenous groups provides a balance between these two objectives.

ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE2019

, , , , ,

We consider the machine teaching problem in a classroom-like setting wherein the teacher has to deliver the same examples to a diverse group of students. Their diversity stems from differences in their initial internal states as well as their learning rates. We prove that a teacher with full knowledge about the learning dynamics of the students can teach a target concept to the entire classroom using O (min {d, N} log 1/epsilon) examples, where d is the ambient dimension of the problem, N is the number of learners, and epsilon is the accuracy parameter. We show the robustness of our teaching strategy when the teacher has limited knowledge of the learners' internal dynamics as provided by a noisy oracle. Further, we study the trade-off between the learners' workload and the teacher's cost in teaching the target concept. Our experiments validate our theoretical results and suggest that appropriately partitioning the classroom into homogenous groups provides a balance between these two objectives.

ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE