English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

We have witnessed the continuing arms race between backdoor attacks and thecorresponding defense strategies on Deep Neural Networks (DNNs). Moststate-of-the-art defenses rely on the statistical sanitization of the "inputs"or "latent DNN representations" to capture trojan behaviour. In this paper, wefirst challenge the robustness of such recently reported defenses byintroducing a novel variant of targeted backdoor attack, called "low-confidencebackdoor attack". We also propose a novel defense technique, called "HaS-Nets".  "Low-confidence backdoor attack" exploits the confidence labels assigned topoisoned training samples by giving low values to hide their presence from thedefender, both during training and inference. We evaluate the attack againstfour state-of-the-art defense methods, viz., STRIP, Gradient-Shaping, Februusand ULP-defense, and achieve Attack Success Rate (ASR) of 99%, 63.73%, 91.2%and 80%, respectively.  We next present "HaS-Nets" to resist backdoor insertion in the network duringtraining, using a reasonably small healing dataset, approximately 2% to 15% offull training data, to heal the network at each iteration. We evaluate it fordifferent datasets - Fashion-MNIST, CIFAR-10, Consumer Complaint and UrbanSound - and network architectures - MLPs, 2D-CNNs, 1D-CNNs. Our experimentsshow that "HaS-Nets" can decrease ASRs from over 90% to less than 15%,independent of the dataset, attack configuration and network architecture.

HaS-Nets: A Heal and Select Mechanism to Defend DNNs Against Backdoor Attacks for Data Collection Scenarios