Fetching the paper…

MalModel: Hiding Malicious Payload in Mobile Deep Learning Models with Black-box Backdoor Attack · Around