Learning scene-specific pedestrian detectors without real data

By Hironori Hattori, Vishnu Naresh Boddeti, Kris M. Kitani, Takeo Kanade. 2015

146 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

The Cityscapes Dataset for Semantic Urban Scene Understanding (2016, background): “Overall, we believe that the unique characteristics of our dataset (e.g. scale variation, amount of small objects, focus on urban street scenes) allow for more such novel insights.”
Recognition in Terra Incognita (2018, methodology): “[48,49,50,51,52] look at methods for adapting detectors from one data source to another, such as from synthetic to real data or from images to video.”
Domain Adaptation for Visual Applications: A Comprehensive Survey (2017, background): “viewpoint estimation [161, 162, 163, 164], object and object part detection [165, 166, 167, 168, 169, 170, 171, 172], segmentation and scene understanding [173, 174, 175].”
Domain Adaptive Faster R-CNN for Object Detection in the Wild (2018, background): “There also exists work to learn detectors from alternative sources, such as from images to videos [54], from 3D models [46, 53], or from synthetic models [25].”
Conditional Generative Adversarial Network for Structured Domain Adaptation (2018, background): “Recent advances in computer graphics have stimulated the interests to train deep learning models on photo-realistic synthetic data with computergenerated annotations, and apply them in real-world settings [12, 29, 47, 37, 41, 48, 20, 19, 32, 22, 34, 33].”
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], singleview reconstruction [58], view-point estimation [84, 119], 2D human pose estimation [93,96,105], 3D human pose estimation [18, 27, 39, 104”
Geometric Image Synthesis (2018, background): “Several works consider the augmentation of images with virtual [5,15] or scanned humans [4,34].”
Expecting the Unexpected: Training Detectors for Unusual Pedestrians with Adversarial Imposters (2017, methodology): “Synthetic training data is an actively explored topic because it provides a potentially inﬁnite well of annotated data for training data-hungry architectures [24, 30, 14, 19, 40, 42].”
The SYNTHIA Dataset: A Large Collection of Synthetic Images for Semantic Segmentation of Urban Scenes (2016, methodology): “…section 2, the use of visually realistic synthetic images is gaining attention in recent years (e.g., training in virtual words [21] [34] [1] [26] [12], synthesizing images with real-world backgrounds and inserted virtual objects [28] [27]) due to the possibility of having diversified samples…”
A Benchmark and Simulator for UAV Tracking (2016, background): “pedestrians [14], 3D scenes [31], and 2D/3D objects [15,32]), where a high demand for annotated data exists.”
Synthetic Data for Deep Learning (2019, methodology): “[236] trained scene-specific pedestrian detectors on a purely synthetic dataset, superimposing rendered pedestrians onto a fixed real scene background; synthetic data has also been used for pedestrian detection by Marin et al.”
Structured Domain Randomization: Bridging the Reality Gap by Context-Aware Synthetic Data (2018, background): “Other researchers [25], [26] have also found context to be important.”
UnrealCV: Connecting Computer Vision to Unreal Engine (2016, background): “Researchers have shown that training on synthetic images is helpful for real world tasks[18,16,11,14,21].”
How Useful Is Photo-Realistic Rendering for Visual Learning? (2016, background): “[18] showed that deep networks are not invariant to certain appearance changes, and use rendered data to augment the training data, and in [9,25] rendered data is used to train pedestrian detectors.”
Driving in the Matrix: Can virtual worlds replace human-generated annotations for real world tasks? (2016, background): “Further work inserted synthetic pedestrians at a variety of angles into real scenes to train a part-based model classifier [14].”
MoCap-guided Data Augmentation for 3D Pose Estimation in the Wild (2016, background): “In [11], a scene-specific pedestrian detectors was learned without real data while [9] synthesized virtual samples with a generative model to enhance the classification performance of a discriminative model.”
Augmented Reality Meets Computer Vision: Efficient Data Generation for Urban Driving Scenes (2017, unknown): “[12] present a scene-specific pedestrian detector using only synthetic data.”