Real-Time Pedestrian Detection with Deep Network Cascades

By Anelia Angelova, Alex Krizhevsky, Vincent Vanhoucke, Abhijit S. Ogale, Dave Ferguson. 2015

253 citations recorded by the source corpus.

We present a new real-time approach to object detection that exploits the efficiency of cascade classifiers with the accuracy of deep neural networks. Deep networks have been shown to excel at classification tasks, and their ability to operate on raw pixel input without the need to design special features is very appealing. However, deep nets are notoriously slow at inference time. In this paper, we propose an approach that cascades deep nets and fast features, that is both very fast and very accurate. We apply it to the challenging task of pedestrian detection. Our algorithm runs in real-time at 15 frames per second. The resulting approach achieves a 26.2% average miss rate on the Caltech Pedestrian detection benchmark, which is competitive with the very best reported results. It is the first work we are aware of that achieves very high accuracy while running in real-time.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Dynamic Neural Networks: A Survey (2021, background): “Image Sa Object detection (face [40], [203], [204], facial point [205], pedestrian [206], general [33], [207], [208], [209], [210]) Image segmentation [107], [211], [212], Super resolution [213], Style transfer [214], Coarse-to-fine classification [215]”
A Survey of Deep Network Solutions for Learning Control in Robotics: From Reinforcement to Imitation (2016, neutral): “. The model, trained end-to-end, can output both the bounding box and the class of the objects from the single image directly. The same strategy was used for pedestrian detection for autonomous cars (Angelova et al., 2015a; Chen et al., 2016c; Angelova et al., 2015b). Besides RGB images, RGB-D data ”
TensorFuzz: Debugging Neural Networks with Coverage-Guided Fuzzing (2018, background): “Neural networks are gradually becoming used in more contexts that affect human lives, including for medical diagnosis [13], in autonomous vehicles [2, 7, 19], as input into corporate and judicial decision making processes [4, 37], in air trafﬁc control [23], and in power grid control [40].”
VisualBackProp: visualizing CNNs for autonomous driving (2016, background): “They also demonstrate excellent performance in pedestrian detection problems [5] and more recently were successfully used in car-driving systems [6, 7].”
Recent Advances in Deep Learning for Object Detection (2019, background): “Deep learning based methods for pedestrian detection [8, 9, 10, 205, 206, 207, 208, 209, 210, 211] showed excellent performance and achieved state-of-the-art results on public benchmarks.”
Bi-box Regression for Pedestrian Detection and Occlusion Estimation (2018, methodology): “A cascade of deep CNNs are proposed in [1] to achieve real-time pedestrian detection by first using tiny deep CNNs to reject a large number of negative proposals and then using large deep CNNs to classify remaining proposals.”
Survey on Emotional Body Gesture Recognition (2018, background): “were passed through a deep network obtaining in this way a trade-off between speed and accuracy [60].”
InferLine: latency-aware provisioning and scaling for prediction serving pipelines (2020, methodology): “This conditional evaluation pattern appears in bandit algorithms [3, 20] used for model personalization as well as more general cascaded prediction pipelines [2, 14, 24, 34].”
Deep-learning in Mobile Robotics - from Perception to Control Systems: A Survey on Why and Why not (2016, methodology): “The same strategy was used for pedestrian detection for autonomous cars (Angelova et al., 2015a; Chen et al., 2016c; Angelova et al., 2015b).”
Pedestrian Detection: The Elephant In The Room (2020, background): “The same trend is also true for the pedestrian detection [2,17,7].”
CityPersons: A Diverse Dataset for Pedestrian Detection (2017, background): “Initial attempts to apply convnets for pedestrian detection, used existing detectors (mainly decision forests over hand-crafted features [2, 33]) outputs and re-scored them with a convnet classifier (plus bounding box regression) [16, 28, 1, 27].”
VNect (2017, methodology): “The use of CNNs in real time has been explored for variants of the object detection problem, for instance bounding box detection and pedestrian detection methods have leveraged application specific architectures [Angelova et al. 2015; Liu et al. 2016; Redmon et al. 2015] and preprocessing steps [Ren”
IDK Cascades: Fast Deep Learning by Learning not to Overthink (2017, background): “[6] proposed using deep network cascades and achieved real-time performances on pedestrian detection tasks.”
VisualBackProp: efficient visualization of CNNs (2018, background): “tly addressed with CNNs [1], including image recognition [2], speech recognition [3], and natural language processing [4]. They also demonstrate excellent performance in pedestrian detection problems [5] and more recently were successfully used in car-driving systems [6,7]. It therefore becomes incr”
Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers (2016, methodology): “Recently, a few methods [33, 21, 1] are proposed to use cascaded detection framework with CNN, but most of them employ another simpler network to “preprocess” object proposals and use a deeper architecture to evaluate surviving candidates.”
Adaptive Feeding: Achieving Fast and Accurate Detections by Adaptively Combining Object Detectors (2017, background): “Recently, there are also some researches utilizing cascaded and boosting methods [26] [1] [21] [32] [28].”
Computer vision and deep learning techniques for pedestrian detection and tracking: A survey (2018, background): “In [205] , Angelova et al. presented a new real-time approach to object detection that exploits the eﬃciency of cascade classiﬁers with the accuracy of deep neural networks.”
Occlusion-aware R-CNN: Detecting Pedestrians in a Crowd (2018, background): “[1] combine the ideas of fast cascade and a deep network to detect pedestrian.”
Hard-Aware Deeply Cascaded Embedding (2016, methodology): “Our method of ensembling a set of models of different complexities in a cascaded manner shares the same spirit as the acceleration technique used in object/face detection [1, 16, 30, 42].”