ROSAR: An Adversarial Re-Training Framework for Robust Side-Scan Sonar Object Detection
This paper introduces ROSAR, a novel framework enhancing the robustness of deep learning object detection models tailored for side-scan sonar (SSS) images, generated by autonomous underwater vehicles using sonar sensors. By extending our prior work on knowledge distillation (KD), this framework integrates KD with adversarial retraining to address the dual challenges of model efficiency and robustness against SSS noises. We introduce three novel, publicly available SSS datasets, capturing different sonar setups and noise conditions. We propose and formalize two SSS safety properties and utilize them to generate adversarial datasets for retraining. Through a comparative analysis of projected gradient descent (PGD) and patch-based adversarial attacks, ROSAR demonstrates significant improvements in model robustness and detection accuracy under SSS-specific conditions, enhancing the model's robustness by up to 1.85%. ROSAR is available at https://github.com/remaro-network/ROSAR-framework.
Code (1)
Tasks
Knowledge Distillationobject-detectionObject DetectionMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Strategies for targeting chondrosarcomas in vivo and molecular dissection of oncogenic events in chondrosarcomas: is epigenetics the culprit?
It is obvious that both epigenetic and non-epigenetic actors contribute to tumorigenesis in chondrosarcomas and more generally in other cancers. Thus, the main altered pathways in chondrosarcomas are now well established…
ZeroSARAH: Efficient Nonconvex Finite-Sum Optimization with Zero Full Gradient Computation
We propose ZeroSARAH -- a novel variant of the variance-reduced method SARAH (Nguyen et al., 2017) -- for minimizing the average of a large number of nonconvex functions $\frac{1}{n}\sum_{i=1}^{n}f_i(x)$. To the best of …
Federated LearningZeroSARAH: Efficient Nonconvex Finite-Sum Optimization with Zero Full Gradient Computations
We propose ZeroSARAH -- a novel variant of the variance-reduced method SARAH (Nguyen et al., 2017) -- for minimizing the average of a large number of nonconvex functions $\frac{1}{n}\sum_{i=1}^{n}f_i(x)$. To the best of …
Federated LearningProSarc: Prosody-Aware Sarcasm Recognition Framework via Temporal Prosodic Incongruity
We present ProSarc, an audio-only framework that detects sarcasm by modelling temporal prosodic incongruity, that is, the mismatch between local prosodic dynamics and the utterance-level emotional baseline. Dual encoding…
Efficient Rotating Synthetic Aperture Radar Imaging via Robust Sparse Array Synthesis
Rotating Synthetic Aperture Radar (ROSAR) can generate a 360$^\circ$ image of its surrounding environment using the collected data from a single moving track. Due to its non-linear track, the Back-Projection Algorithm (B…