Scale optimization for full-image-CNN vehicle detection

Abstract

Many state-of-the-art general object detection methods make use of shared full-image convolutional features (as in Faster R-CNN). This achieves a reasonable test-phase computation time while enjoys the discriminative power provided by large Convolutional Neural Network (CNN) models. Such designs excel on benchmarks 1 which contain natural images but which have very unnatural distributions, i.e. they have an unnaturally high-frequency of the target classes and a bias towards a "friendly" or "dominant" object scale. In this paper we present further study of the use and adaptation of the Faster R-CNN object detection method for datasets presenting natural scale distribution and unbiased real-world object frequency. In particular, we show that better alignment of the detector scale sensitivity to the extant distribution improves vehicle detection performance. We do this by modifying both the selection of Region Proposals, and through using more scaleappropriate full-image convolution features within the CNN model. By selecting better scales in the region proposal input and by combining feature maps through careful design of the convolutional neural network, we improve performance on smaller objects. We significantly increase detection AP for the KITTI dataset car class from 76.3% on our baseline Faster R-CNN detector to 83.6% in our improved detector.

Towards an End-to-End Framework of CCTV-Based Urban Traffic Volume Detection and Prediction

“…They found that a pre-trained Faster R-CNN ResNet model fine-tuned with 700 CCTV images could identify more small vehicles in the image background, providing better precision and fewer false negatives than the fine-tuned SSD model. To address the issue of varying vehicle scales in a natural scene, the study in [ 32 ] optimized the default anchor box size and changed the default combination of convolutional layers in the Faster R-CNN architecture using KITTI image datasets. That way, they further improved its performance in detecting small vehicles by circa 7%.…”

Analysis of the Influence of Foggy Weather Environment on the Detection Effect of Machine Vision Obstacles

“…Ye Wang [11] optimized anchor generation and improved performance through region of interest allocation, making the number of features after pooling more suitable for final prediction. Yang Gao et al [12] improved the convolution layer region scheme of the Faster R-CNN model and improved the detection accuracy by 7.3% through the KITTI data test. Reference [13] presents a pre-processing pipeline on Faster R-CNN to improve the training and detection speed.…”

Geospatial Object Detection in High Resolution Satellite Images Based on Multi-Scale Convolutional Neural Network

“…2018, 10, 131 2 of 21 (CNN), several design variations using region based CNN have generated the state-of-the-art performance against traditional multi-class object detection benchmarks [16][17][18][19][20]. These benchmark datasets typically present target objects with "friendly" or dominant scales because those images in a large pool of available images and objects with significant scales, could be more easily selected [21]. Unlike objects on these benchmark datasets, objects on HRS images are much smaller, including fixed shape objects (e.g., airplanes, ships, vehicles, etc.)…”