You Only Look Once: Unified, Real-Time Object Detection
By Joseph Redmon, Santosh Kumar Divvala, Ross B. Girshick, Ali Farhadi. 2015
46,861 citations recorded by the source corpus.
We present YOLO, a new approach to object detection. Prior work on object detection repurposes classifiers to perform detection. Instead, we frame object detection as a regression problem to spatially separated bounding boxes and associated class probabilities. A single neural network predicts bounding boxes and class probabilities directly from full images in one evaluation. Since the whole detection pipeline is a single network, it can be optimized end-to-end directly on detection performance. Our unified architecture is extremely fast. Our base YOLO model processes images in real-time at 45 frames per second. A smaller version of the network, Fast YOLO, processes an astounding 155 frames per second while still achieving double the mAP of other real-time detectors. Compared to state-of-the-art detection systems, YOLO makes more localization errors but is less likely to predict false positives on background. Finally, YOLO learns very general representations of objects. It outperforms other detection methods, including DPM and R-CNN, when generalizing from natural images to other domains like artwork.
Subfields: 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 Computer Vision and Pattern Recognition
- Luc Van Gool
- Dacheng Tao
- Xiaogang Wang
- Ming-Hsuan Yang
- Trevor Darrell
- Jitendra Malik
- Andrew Zisserman
- Yi Yang
Citations in the corpus
- End-to-End Object Detection with Transformers (2020, methodology): “Several object detectors [9,25,35] used the bipartite matching loss.”
- Rethinking ImageNet Pre-Training (2018, background): “As modern object detectors [9, 15, 8, 36, 35, 28, 26, 13] evolved under the pre-training paradigm, the belief that training from scratch is non-trivial became conventional wisdom.”
- Objects as Points (2019, background): “Current object detectors represent each object through an axis-aligned bounding box that tightly encompasses the object [18, 19, 33, 43, 46].”
- Elucidating image-to-set prediction: An analysis of models, losses and datasets (2019, methodology): “Major advances in image understanding tasks [57, 18, 50, 49, 37, 23, 17] have been enabled by the introduction of large scale datasets such as ImageNet [53], MSCOCO [30] or Cityscapes [10].”
- Generalized Intersection Over Union: A Metric and a Loss for Bounding Box Regression (2019, methodology): “in YOLO v1[19] propose a direct regression on the bounding box parameters with a small tweak to predict square root of the bounding box size to remedy scale sensitivity.”
- Deep Perm-Set Net: Learn to predict sets with unknown permutation and cardinality using deep neural networks (2018, methodology): “Existing approaches formulate this problem using as a pre-defined and fixed-sized grid Redmon et al. (2016); Redmon & Farhadi 1Throughout the paper, we use the term structured data for vectors, matrices and generally tensors which are fixed in size and ordered, i.e. any permutation of its values…”
- Learning Non-maximum Suppression (2017, background): “This last stage is commonly referred to as “non-maximum suppression” (NMS) [10, 9, 21, 7, 20, 16].”
- Soft-NMS — Improving Object Detection with One Line of Code (2017, methodology): “For detectors like SSD [18] and YOLOv2 [23] which are not proposal based, with the linear function, Soft-NMS only obtains an improvement of 0.”
- Learning to decompose for object detection and instance segmentation (2015, background): “Another tack, e.g., Redmon et al. (2015), removes the region proposal step and directly predicts multiple objects and their locations with a single CNN evaluation.”
- Speed/Accuracy Trade-Offs for Modern Convolutional Object Detectors (2016, methodology): “Unfortunately, only a small subset of papers (e.g., R-FCN [6], SSD [22] YOLO [25]) discuss running time in any detail.”
- R-FCN: Object Detection via Region-based Fully Convolutional Networks (2016, background): “Although methods that do not rely on region proposal do exist ( e.g. , [18, 15]), region-based systems still possess leading accuracy on several benchmarks [5, 14, 21].”