Unsupervised Hard Example Mining from Videos for Improved Object Detection
This work addresses the challenge of obtaining hard examples for training object detectors, which is crucial for improving accuracy in applications like face and pedestrian detection, though it is incremental as it builds on existing hard example mining techniques.
The paper tackles the problem of sparse and expensive hard negative examples in object detection by automatically mining them from unlabeled video sequences using temporal isolation criteria, resulting in significant performance improvements across multiple architectures and datasets.
Important gains have recently been obtained in object detection by using training objectives that focus on {\em hard negative} examples, i.e., negative examples that are currently rated as positive or ambiguous by the detector. These examples can strongly influence parameters when the network is trained to correct them. Unfortunately, they are often sparse in the training data, and are expensive to obtain. In this work, we show how large numbers of hard negatives can be obtained {\em automatically} by analyzing the output of a trained detector on video sequences. In particular, detections that are {\em isolated in time}, i.e., that have no associated preceding or following detections, are likely to be hard negatives. We describe simple procedures for mining large numbers of such hard negatives (and also hard {\em positives}) from unlabeled video data. Our experiments show that retraining detectors on these automatically obtained examples often significantly improves performance. We present experiments on multiple architectures and multiple data sets, including face detection, pedestrian detection and other object categories.