Environmental conditions such as low-light at night, fog scattering, glare artifacts, rain streaks, and rain smear distortions are significant issues of camera-based perception in Autonomous Vehicles (AVs). These degradations alter the statistics of the scene, mask structure, introduce non-uniform noise, and adversely affect downstream vision processes, including detection and tracking. To overcome this shortcoming, this paper presents a lightweight TinyVGG-based degradation classification system that runs in real time. The network extracts discriminative spatial features with hierarchical convolutional encoding and projects them to a lower-dimensional semantic representation with fully connected layers and a multi-class predictor based on SoftMax. In addition, a confidence estimation mechanism is incorporated to assess the reliability of each prediction, providing an additional level of robustness and interpretability, particularly in safety-critical driving scenarios. A custom real-world driving dataset was built and annotated into five degradation categories. The evaluation shows high performance, with an overall accuracy of 98.17%, a macro F1-score of 98.20%, and an ROC-AUC of 0.9978, along with low computational overhead and real-time inference. These results support the usefulness and feasibility of the proposed framework in embedded automotive systems with strict latency and resource limitations.
Abbas et al. (Tue,) studied this question.