Pedestrian detection for driving assistance systems: single-frame classification and system level performance

By Amnon Shashua, Yoram Gdalyahu, Gaby Hayun. IEEE Intelligent Vehicles Symposium, 2004

419 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Signal Processing

View Paper: https://www.semanticscholar.org/paper/37a15ce03c26ec83d95bf4aaf756a41370d50353

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

In Signal Processing

Citations in the corpus

Object scene flow for autonomous vehicles (2015, background): “Important applications include mobile robotics and autonomous driving where 3D object motion is a fundamental input to high-level tasks such as scene understanding, obstacle avoidance or path planning [5,7,10,28].”
Multi-cue Pedestrian Detection and Tracking from a Moving Vehicle (2007, methodology): “Very recently, additional attempts have been made towards reducing classification complexity by manually separating the pedestrian training set in nonoverlapping sub-sets (i.e. based on pedestrian heading direction) Shashua et al. (2004), Grubb et al. (2004), and Shimizu and Poggio (2004).”
Unsupervised Bayesian Detection of Independent Motion in Crowds (2006, background): “They, like [13, 20, 4], and [27] for the static case, have a probabilistic scheme for finding body part primitives that are likely to fit together.”
Crowd analysis: a survey (2008, background): “[81] describe a functional and architectural breakdown pedestrian detection system.”
Pedestrian detection: A benchmark (2009, background): “Automotive applications [12, 14, 34] are particularly compelling as they have the potential to save numerous lives [39].”
Detection and Tracking of Multiple, Partially Occluded Humans by Bayesian Combination of Edgelet based Part Detectors (2007, neutral): “The methods of Shashua et al. (2004) and Mikolajczyk et al. (2004) both achieved better results than that of Mohan et al. (2001), but there is no direct comparison between (Shashua et al., 2004) and (Mikolajczyk et al., 2004).”
Survey of Pedestrian Detection for Advanced Driver Assistance Systems (2010, methodology): “In [97], Shashua et al. propose a multiframe approval process that consists of validating the pedestrian-classified ROIs by collecting information from several frames: gait pattern, inward motion, confidence of the single-frame classification, etc.”
Co-occurrence Histograms of Oriented Gradients for Pedestrian Detection (2009, methodology): “Since shapelets are the combination of edges, they can express more detailed shape information than what SIFT/HOG feature descriptors can.”
Human detection from images and videos: A survey (2016, methodology): “Its performance was also verified in the work of Mikolajczyk and Schmid [101] and used to encode the parts of the object in [102, 140].”
Computer Vision for Autonomous Vehicles: Problems, Datasets and State-of-the-Art (2017, background): “Temporal Cues: Similarly, Shashua et al. (2004) point out the importance of good features for the person detection task.”
Virtual to Real Adaptation of Pedestrian Detectors (2020, background): “They range from video surveillance [1,2], robotics, automotive [3,4] and assistive technologies to people with visual disabilities [5], just to name a few.”
Multi-cue onboard pedestrian detection (2009, methodology): “Sashua et al. [ 26 ] use parts and employ histograms of gradients as features.”
An Experimental Study on Pedestrian Classification (2006, background): “Shashua et al. [8], for instance, extract a feature vector from each of nine fixed subregions.”
Pedestrian Detection: An Evaluation of the State of the Art (2012, background): “In earlier work, Shashua et al. [48] proposed a similar representation for characterizing spatially localized parts for modeling pedestrians.”
New features and insights for pedestrian detection (2010, background): “All competitive classifiers we know of employ statistical learning techniques to learn the mapping from features to scores (indicating the likelihood of a pedestrian being present)—usually either support vector machines [3, 13, 17, 19, 27] or some variant of boosting [23, 27, 28, 30].”
Pedestrian Protection Systems: Issues, Survey, and Challenges (2007, background): “patterns [79] are some of the cues used to detect and verify pedestrian candidates.”