Darknet convolutional groups=0 SIGFPE PoC
This repository contains a benign security research PoC for a Darknet config
that crashes OpenCV DNN with SIGFPE.
Files:
control_same_size.cfgmalicious_groups_zero.cfgtiny_shared.weightsreproduce.py
Observed behavior:
- control config:
- loads successfully
- malicious config:
- with
[convolutional] groups=0 cv2.dnn.readNetFromDarknet(...)terminates with signal8
- with
Reproduction:
python3 build_poc.py
python3 reproduce.py
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support