Machine Unlearning: A Survey

Machine learning has attracted widespread attention and evolved into an enabling technology for a wide range of highly successful applications, such as intelligent computer vision, speech recognition, medical diagnosis, and more. Yet a special need has arisen where, due to privacy, usability, and/or the right to be forgotten, information about some specific samples needs to be removed from a model, called machine unlearning. This emerging technology has drawn significant interest from both academics and industry due to its innovation and practicality. At the same time, this ambitious problem has led to numerous research efforts aimed at confronting its challenges. To the best of our knowledge, no study has analyzed this complex topic or compared the feasibility of existing unlearning solutions in different kinds of scenarios. Accordingly, with this survey, we aim to capture the key concepts of unlearning techniques. The existing solutions are classified and summarized based on their characteristics within an up-to-date and comprehensive review of each category's advantages and limitations. The survey concludes by highlighting some of the outstanding issues with unlearning techniques, along with some feasible directions for new research opportunities.

Neuro-Modulated Hebbian Learning for Fully Test-Time Adaptation

Fully test-time adaptation aims to adapt the network model based on sequential analysis of input samples during the inference stage to address the cross-domain performance degradation problem of deep neural networks. We take inspiration from the biological plausibility learning where the neuron responses are tuned based on a local synapse-change procedure and activated by competitive lateral inhibition rules. Based on these feed-forward learning rules, we design a soft Hebbian learning process which provides an unsupervised and effective mechanism for online adaptation. We observe that the performance of this feed-forward Hebbian learning for fully test-time adaptation can be significantly improved by incorporating a feedback neuro-modulation layer. It is able to fine-tune the neuron responses based on the external feedback generated by the error back-propagation from the top inference layers. This leads to our proposed neuro-modulated Hebbian learning (NHL) method for fully test-time adaptation. With the unsupervised feed-forward soft Hebbian learning being combined with a learned neuro-modulator to capture feedback from external responses, the source model can be effectively adapted during the testing process. Experimental results on benchmark datasets demonstrate that our proposed method can significantly improve the adaptation performance of network models and outperforms existing state-of-the-art methods.

A Promising Method for Touch-typing Keyboard Rendering

We have developed a novel button click rendering mechanism based on active lateral force feedback. The effect can be localized because electroadhesion between a finger and a surface can be localized. Psychophysical experiments were conducted to evaluate the quality of a rendered button click, which subjects judged to be acceptable. Both the experiment results and the subjects' comments confirm that this button click rendering mechanism has the ability to generate a range of realistic button click sensations that could match subjects' different preferences. We can thus generate a button click on a flat surface without macroscopic motion of the surface in the lateral or normal direction, and we can localize this haptic effect to an individual finger. This mechanism is promising for touch-typing keyboard rendering.

How the Mechanical Properties and Thickness of Glass Affect TPaD Performance

One well-known class of surface haptic devices that we have called TPaDs (Tactile Pattern Displays) uses ultrasonic transverse vibrations of a touch surface to modulate fingertip friction. This paper addresses the power consumption of glass TPaDs, which is an important consideration in the context of mobile touchscreens. In particular, based on existing ultrasonic friction reduction models, we consider how the mechanical properties (density and Young's modulus) and thickness of commonly-used glass formulations affect TPaD performance, namely the relation between friction reduction ability and real power consumption. Experiments performed with eight types of TPaDs and an electromechanical model for the fingertip-TPaD system indicate: 1) TPaD performance decreases as glass thickness increases; 2) TPaD performance increases as the Young's modulus and density of glass decrease; 3) counterintuitively, real power consumption of a TPaD decreases as the contact force increases. Proper applications of these results can lead to significant increases in TPaD performance.