Multispectral Pedestrian Detection using Deep Fusion Convolutional Neural Networks

By Jörg Wagner, Volker Fischer, Michael Herman, Sven Behnke. The European Symposium on Artificial Neural Networks, 2016

298 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/837ab5c420010ab850a800d39605cbaafd8e0203

How researchers see this paper

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

1. Di Feng
2. Kailai Zhou
3. Chengyang Li
4. Qingyun Fang
5. Linsen Chen
6. Jamil Fayyad
7. Élisa Fromont
8. Dan Song

Citations in the corpus

• Multispectral Deep Neural Networks for Pedestrian Detection (2016, background): “Nevertheless, except very recent effort [12, 42], most of previous studies focus on detecting pedestrians with color or thermal images only, rather than leveraging color and thermal images simultaneously.”
• Fusion of Multispectral Data Through Illumination-aware Deep Neural Networks for Pedestrian Detection (2018, background): “[35] presented the first application of DNNs for multispectral pedestrian d etection and evaluated the performance of two decision netwo rks (early-fusion and late-fusion).”
• Automatic autonomous vision-based power line inspection: A review of current status and the potential role of deep learning (2018, methodology): “drawbacks, such as their sensitivity to illumination changes [91] and their incapability of detecting faults that are invisible to the unaided human eye (e.”
• A Systematic Review of Perception System and Simulators for Autonomous Vehicles Research (2019, methodology): “In [37], multispectral information is used from a thermal camera and an RGB camera (TRGB channels) and a multi-layer fused CNN to detect pedestrians under adverse illumination conditions.”
• Illumination-aware Faster R-CNN for Robust Multispectral Pedestrian Detection (2018, methodology): “* Results of [13, 14] on improved test annotations are unavailable since we cannot obtain their codes or detections.”
• Improving Multispectral Pedestrian Detection by Addressing Modality Imbalance Problems (2020, methodology): “e cornerstone we should consider in multispectral pedestrian detection. Unfortunately, existing RGB-Thermal based detection methods simply fuse the RGB and the thermal input/features by concatenation [23,42,22,37]. The inherent complementary is not fully exploited yet between dierent modalities. To ”
• Deep Learning Sensor Fusion for Autonomous Vehicle Perception and Localization: A Review (2020, background): “[39] studied the impact of fusing thermal images and visible images in pedestrian detection in both daytime and ighttime.”
• Multispectral Fusion for Object Detection with Cyclic Fuse-and-Refine Blocks (2020, background): “The first study on CNN-based multispectral pedestrian detection is made by [3], and they evaluate two fusion strategies: early and late fusions.”
• Deep Multi-Modal Object Detection and Semantic Segmentation for Autonomous Driving: Datasets, Methods, and Challenges (2019, background): “[105], [107], [108], [110], [116]–[119], [122] deal with the 2D object detection problem on the front-view 2D image plane.”
• Cross-Modality Fusion Transformer for Multispectral Object Detection (2021, background): “The first study of this type [20] investigates two deep fusion architectures (early fusion and late fusion) 40”