The anomaly detection in photovoltaic (PV) cell electroluminescence (EL) image is of great significance for the vision-based fault diagnosis. Many researchers are committed to solving this problem, but a large-scale open-world dataset is required to validate their novel ideas. We build a PV EL Anomaly Detection (PVEL-AD ^{1, 2, 3} ) dataset for polycrystalline solar cell, which contains 36 543 near-infrared images with various internal defects and heterogeneous background. This dataset contains anomaly free images and anomalous images with ten different categories. Moreover, 37 380 ground truth bounding boxes are provided for eight types of defects. We also carry out a comprehensive evaluation of the state-of-the-art object detection methods based on deep learning. The evaluation results on this dataset provide the initial benchmark, which is convenient for follow-up researchers to conduct experimental comparisons. To the best of our knowledge, this is the first public dataset for PV solar cell anomaly detection that provides box-wise ground truth. Furthermore, this dataset can also be used for the evaluation of many computer vision tasks such as few-shot detection, one-class classification, and anomaly generation.

The multiscale defect detection for photovoltaic (PV) cell electroluminescence (EL) images is a challenging task, due to the feature vanishing as network deepens. To address this problem, an attention-based top-down and bottom-up architecture is developed to accomplish multiscale feature fusion. This architecture, called bidirectional attention feature pyramid network (BAFPN), can make all layers of the pyramid share similar semantic features. In BAFPN, cosine similarity is employed to measure the importance of each pixel in the fused features. Furthermore, a novel object detector is proposed, called BAF-Detector, which embeds BAFPN into region proposal network in Faster RCNN+FPN. BAFPN improves the robustness of the network to scales, thus the proposed detector achieves a good performance in multiscale defects detection task. Finally, the experimental results on a large-scale EL dataset, including 3629 images, 2129 of which are defective, show that the proposed method achieves 98.70% (F-measure), 88.07% (mAP), and 73.29% (IoU) in terms of multiscale defects classification and detection results in raw PV cell EL images.

Context modeling or multi-level feature fusion methods have been proved to be effective in improving semantic segmentation performance. However, they are not specialized to deal with the problems of pixel-context mismatch and spatial feature misalignment, and the high computational complexity hinders their widespread application in real-time scenarios. In this work, we propose a lightweight Context and Spatial Feature Calibration Network (CSFCN) to address the above issues with pooling-based and sampling-based attention mechanisms. CSFCN contains two core modules: Context Feature Calibration (CFC) module and Spatial Feature Calibration (SFC) module. CFC adopts a cascaded pyramid pooling module to efficiently capture nested contexts, and then aggregates private contexts for each pixel based on pixel-context similarity to realize context feature calibration. SFC splits features into multiple groups of sub-features along the channel dimension and propagates sub-features therein by the learnable sampling to achieve spatial feature calibration. Extensive experiments on the Cityscapes and CamVid datasets illustrate that our method achieves a state-of-the-art trade-off between speed and accuracy. Concretely, our method achieves 78.7% mIoU with 70.0 FPS and 77.8% mIoU with 179.2 FPS on the Cityscapes and CamVid test sets, respectively.

For the problem of action detection, most existing methods require that relevant portions of the action of interest in training videos have been manually annotated with bounding boxes. Some recent works tried to avoid tedious manual annotation , and proposed to automatically identify the relevant portions in training videos. However, these methods only concerned the identification in either spatial or temporal domain, and may get irrelevant contents from another domain. These irrelevant contents are usually undesirable in the training phase, which will lead to a degradation of the detection performance. This paper advances prior work by proposing a joint learning framework to simultaneously identify the spatial and temporal extents of the action of interest in training videos. To get pixel-level localization results, our method uses dense trajectories extracted from videos as local features to represent actions. We first present a trajectory split-and-merge algorithm to segment a video into the background and several separated foreground moving objects. In this algorithm, the inherent temporal smoothness of human actions is exploited to facilitate segmentation. Then, with the latent SVM framework on segmentation results, spatial and temporal extents of the action of interest are treated as latent variables that are inferred simultaneously with action recognition. Experiments on two challenging datasets show that action detection with our learned spatial and temporal extents is superior than state-of-the-art methods.