Cascade R-CNN: Delving Into High Quality Object Detection

By Zhaowei Cai, Nuno Vasconcelos. 2017

6,288 citations recorded by the source corpus.

In object detection, an intersection over union (IoU) threshold is required to define positives and negatives. An object detector, trained with low IoU threshold, e.g. 0.5, usually produces noisy detections. However, detection performance tends to degrade with increasing the IoU thresholds. Two main factors are responsible for this: 1) overfitting during training, due to exponentially vanishing positive samples, and 2) inference-time mismatch between the IoUs for which the detector is optimal and those of the input hypotheses. A multi-stage object detection architecture, the Cascade R-CNN, is proposed to address these problems. It consists of a sequence of detectors trained with increasing IoU thresholds, to be sequentially more selective against close false positives. The detectors are trained stage by stage, leveraging the observation that the output of a detector is a good distribution for training the next higher quality detector. The resampling of progressively improved hypotheses guarantees that all detectors have a positive set of examples of equivalent size, reducing the overfitting problem. The same cascade procedure is applied at inference, enabling a closer match between the hypotheses and the detector quality of each stage. A simple implementation of the Cascade R-CNN is shown to surpass all single-model object detectors on the challenging COCO dataset. Experiments also show that the Cascade R-CNN is widely applicable across detector architectures, achieving consistent gains independently of the baseline detector strength. The code is available at https://github.com/zhaoweicai/cascade-rcnn.

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

Rethinking ImageNet Pre-Training (2018, methodology): “with pre-training on various systems using Mask R-CNN, including: (i) baselines using FPN and GN, (ii) baselines with trainingtime multi-scale augmentation, (iii) baselines with Cascade RCNN [3] and training-time augmentation, and (iv) plus test-time multi-scale augmentation.”
Benchmarking Detection Transfer Learning with Vision Transformers (2021, background): “Mask R-CNN is the foundation of higher complexity/higher performing systems, such as Cascade R-CNN [4] and HTC/HTC++ [6, 29], which may improve upon the results presented here at the cost of additional complexity that is orthogonal to the goal of benchmarking transfer learning.”
Bridging the Gap Between Anchor-Based and Anchor-Free Detection via Adaptive Training Sample Selection (2019, methodology): “The second stage of Grid R-CNN [39] locates objects via predicting grid points with the position sensitive merits of FCN and then determining the bounding box guided by the grid.”
Cascade R-CNN: High Quality Object Detection and Instance Segmentation (2019, methodology): “instance segmentation literatures, some of which were introduced after [3], e.”
Segment Anything (2023, methodology): “We compare to a strong baseline implemented as a ViTDet [62] detector (with cascade Mask R-CNN [48, 11] ViT-H).”
Pix2seq: A Language Modeling Framework for Object Detection (2021, methodology): “To predict a set of bounding boxes, architectures of modern detectors are speciﬁcally designed to produce a large set of proposals (Girshick, 2015; Ren et al., 2015; Cai & Vasconcelos, 2018), anchors (Lin et al., 2017b), or window centers (Tian et al., 2019; Zhou et al., 2019).”
Global Pooling, More than Meets the Eye: Position Information is Encoded Channel-Wise in CNNs (2021, positive): “We choose the following three recent pedestrian detection models trained on CityPerson (available in [6]): (i) Faster-RCNN [10] (ii) Cascade-RCNN [1] with the HRNet [11] backbone, and (iii) CSP-ResNet-50 [9].”
Swin Transformer: Hierarchical Vision Transformer using Shifted Windows (2021, methodology): “In this section, we ablate important design elements in the proposed Swin Transformer, using ImageNet-1K image classification, Cascade Mask R-CNN on COCO object detection, and UperNet on ADE20K semantic segmentation.”
OneFormer: One Transformer to Rule Universal Image Segmentation (2022, methodology): “Traditional instance segmentation methods [2, 5, 23] are also formulated as mask classifiers, which predict binary masks and a class label for each mask.”
SOCRATES: Introducing Depth in Visual Wildlife Monitoring Using Stereo Vision (2022, methodology): “We set N = 2 for Mask2Former and N = 6 for Cascade Mask R-CNN due to memory constraints.”
Towards Real-World Prohibited Item Detection: A Large-Scale X-ray Benchmark (2021, methodology): “Cascade R-CNN [4] applies the classic cascade architecture to Faster R-CNN [31].”
K-Net: Towards Unified Image Segmentation (2021, methodology): “37 Cascade Mask R-CNN [3] R-50-FPN 36 38.”
Simple Copy-Paste is a Strong Data Augmentation Method for Instance Segmentation (2020, methodology): “Our strongest model uses Cascade R-CNN [2], EfficientNet-B7 as the backbone and NAS-FPN [17] as the feature pyramid with levels from P3 to P7.”
Hybrid Task Cascade for Instance Segmentation (2019, methodology): “We follow the hyperparameter settings in Cascade R-CNN [5].”