Analysing Domain Shift Factors between Videos and Images for Object Detection
By Vicky Kalogeiton, Vittorio Ferrari, Cordelia Schmid. 2015
68 citations recorded by the source corpus.
Object detection is one of the most important challenges in computer vision. Object detectors are usually trained on bounding-boxes from still images. Recently, video has been used as an alternative source of data. Yet, for a given test domain (image or video), the performance of the detector depends on the domain it was trained on. In this paper, we examine the reasons behind this performance gap. We define and evaluate different domain shift factors: spatial location accuracy, appearance diversity, image quality and aspect distribution. We examine the impact of these factors by comparing performance before and after factoring them out. The results show that all four factors affect the performance of the detectors and their combined effect explains nearly the whole performance gap.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Representation & Self-Supervised Learning
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
- Vittorio Ferrari
- Jasper R. R. Uijlings
- Li Fei-Fei
- Trevor Darrell
- Sung Ju Hwang
- Leonid Sigal
- Dim P. Papadopoulos
- Judy Hoffman
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Vittorio Ferrari
- Bernt Schiele
- Jitendra Malik
- Trevor Darrell
- Ming-Hsuan Yang
- Li Fei-Fei
- Dacheng Tao
In Representation & Self-Supervised Learning
- Li Fei-Fei
- Jasper R. R. Uijlings
- Sung Ju Hwang
- Zelun Luo
- Yuliang Zou
- Judy Hoffman
- Dim P. Papadopoulos
- Frank Keller
Citations in the corpus
- Extreme Clicking for Efficient Object Annotation (2017, background): “However, because of the domain adaptation problem, these detectors are still quite weak compared to ones trained on manually annotated still images [34].”
- Y Qualification Main task Y Return N N Y User qualified ? N Feedback Pass ? Pass ? Annotator training Annotating images (2017, background): “However, because of the domain adaptation problem, these detectors are still quite weak compared to ones trained on manually annotated still images [34].”
- Label Efficient Learning of Transferable Representations acrosss Domains and Tasks (2017, background): “Many recent works [60, 24] study the domain shift between images and video in the object detection settings.”
- Expanding object detector's Horizon: Incremental learning framework for object detection in videos (2015, methodology): “DPM: Performance reported in the papers [25, 18] obtained using Deformable Part Model [8].”