Virtual and Real World Adaptation for Pedestrian Detection

By David Vázquez, Antonio M. López, Javier Marín, Daniel Ponsa, David Gerónimo. 2014

194 citations recorded by the source corpus.

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

Unsupervised Domain Adaptation by Backpropagation (2014, background): “ncountered at “test time”. One particularly important example is synthetic or semi-synthetic imagery, which may come in abundance and fully labeled, but which inevitably look different from real data [13,20,23,21]. Learning a discriminative classiﬁer or other predictor in the presence of a shift bet”
Domain-Adversarial Training of Neural Networks (2015, background): “…example is training an image classiﬁer on synthetic or semi-synthetic images, which may come in abundance and be fully labeled, but which inevitably have a distribution that is diﬀerent from real images (Liebelt and Schmid, 2010; Stark et al., 2010; V´azquez et al., 2014; Sun and Saenko, 2014).”
Domain Adaptation for Visual Applications: A Comprehensive Survey (2017, background): “…have great promise for deep learning across a breadth of computer vision problems, including optical ﬂow [148, 24, 156, 147], object trackers [225, 91], depth estimation from RGB [192], object detection [144, 234, 253] semantic segmentation [109, 179, 176] or human actions recognition [211].”
VirtualWorlds as Proxy for Multi-object Tracking Analysis (2016, background): “ settings. First attempts to use synthetic data for training are mainly limited to using rough synthetic models or synthesized real examples (e.g., of pedestrians [14,15]). In contrast, Mar´ın et al. [16,17,18] went further and positively answer the intriguing question whether one can learn appearan”
The SYNTHIA Dataset: A Large Collection of Synthetic Images for Semantic Segmentation of Urban Scenes (2016, background): “in the same spirit than [42] for a HOG-LBP/SVM setting.”
Synthesizing Training Images for Boosting Human 3D Pose Estimation (2016, background): “Park et al. [37] use layering to reconstruct images with different pose, leading to imprecise and poor resolution synthetic images.”
Configurable 3D Scene Synthesis and 2D Image Rendering with Per-pixel Ground Truth Using Stochastic Grammars (2017, background): “Synthetic image datasets have recently been a source of training data for object detection and correspondence matching [26,32,37,83,91,95,113,117,120,150], single-view reconstruction [58], view-point estimation [84, 119], 2D human pose estimation [93, 96, 105], 3D human pose estimation [18, 27, 39, ”
Category Anchor-Guided Unsupervised Domain Adaptation for Semantic Segmentation (2019, methodology): “To overcome this issue, a variety of domain adaptation methods for classiﬁcation [6, 23, 37, 28, 41, 3, 19], detection [38, 16], and segmentation [7, 14, 13, 27, 34, 40, 21, 47] have been proposed.”
Virtual to Real Adaptation of Pedestrian Detectors (2020, background): “[23, 24] focused their attention on the possibility of performing domain adaptation to map virtual features onto real ones.”
From Virtual to Reality: Fast Adaptation of Virtual Object Detectors to Real Domains (2014, background): “Unsupervised adaptation of detectors was explored in [19] for the pedestrian class.”
Multiview random forest of local experts combining RGB and LIDAR data for pedestrian detection (2015, background): “When the value differences are calculated we accept as equal values the ones included in a defined range, this range (defined as ClipTh) allows that small noises (small value changes) do not affect the texture pattern (More details in [32]).”
Co-regularized Alignment for Unsupervised Domain Adaptation (2018, background): “Domain adaptation finds applications in many practical scenarios, including the special case when source domain consists of simulated or synthetic data (for which labels are readily available from the simulator) and target domain consists of real world observations [43, 39, 5].”
ROAD: Reality Oriented Adaptation for Semantic Segmentation of Urban Scenes (2017, background): “Learning Using Synthetic Data: There are also a few works proposed to learn from synthetic data [41, 46, 33, 44, 31, 38].”
Synthetic Data for Deep Learning (2019, methodology): “in adaptation approaches have been validated or subsequently extended to synthetic-to-real domain adaptation. Xu et al. [664] consider the pedestrian detection problem (this work is a continuation of [612], see Section 2.2.2). They adapt detectors trained on virtual datasets with a boosting-based pr”
VisDA: The Visual Domain Adaptation Challenge (2017, unknown): “CAD-PASCAL 833,140 20-pascal 2 NEXAR [1] 55,000 1-car 3 TPAMI14 [48] 5,616 1-pedestrian 2”
Unsupervised Domain Adaptation with a Relaxed Covariate Shift Assumption (2017, methodology): “As source labeled data, we use the CVC-04 virtual-world pedestrian dataset (Vazquez et al. 2014), which consists of 1208 virtual pedestrians and 1220 (a subset of the available 6828) pedestrian-free cropped images.”
Learning Pedestrian Detection from Virtual Worlds (2019, background): “[21, 22] focused their attention on the possibility to perform domain adaptation in order to map virtual features onto real ones.”
Scenic: a language for scenario specification and scene generation (2018, methodology): “Domain adaptation has enabled synthetic data to successfully train models for several other applications including 3D object detection [23, 36], pedestrian detection [40], and semantic image segmentation [33].”