Adaptive Object Detection Using Adjacency and Zoom Prediction

By Yongxi Lu, Tara Javidi, Svetlana Lazebnik. 2015

84 citations recorded by the source corpus.

State-of-the-art object detection systems rely on an accurate set of region proposals. Several recent methods use a neural network architecture to hypothesize promising object locations. While these approaches are computationally efficient, they rely on fixed image regions as anchors for predictions. In this paper we propose to use a search strategy that adaptively directs computational resources to sub-regions likely to contain objects. Compared to methods based on fixed anchor locations, our approach naturally adapts to cases where object instances are sparse and small. Our approach is comparable in terms of accuracy to the state-of-the-art Faster R-CNN approach while using two orders of magnitude fewer anchors on average. Code is publicly available.

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

Citations in the corpus

Recent advances in convolutional neural networks (2015, background): “ios and scales per feature map location. With this multiple scales setting and their matching strategy, SSD is signicantly more accurate than YOLO. With the benets from super-resolution, Lu et al . [248] propose a top-down search strategy to divide a window into sub-windows recursively, in which an ”
AutoFocus: Efficient Multi-Scale Inference (2018, background): “AutoFocus shares the motivation with saliency and reinforcement learning based methods which perform a guided search while processing images [28, 26, 42, 53, 23, 49, 55, 29, 47], but it is designed to predict small objects in coarser scales and they need not be salient.”
F-BRS: Rethinking Backpropagating Refinement for Interactive Segmentation (2020, background): “It is based on the ideas from object detection [20, 7].”
Deep Learning for Generic Object Detection: A Survey (2018, background): “[141] proposed to generate anchor locations by using a recursive search strategy which can adaptively guide computational resources to focus on subregions likely to contain objects.”
Spatial Memory for Context Reasoning in Object Detection (2017, background): “A large portion of the literature [28, 45, 56] focuses on sequential approaches for region proposals (i.”
Recent Advances in Deep Learning for Object Detection (2019, unknown): “proposed AZnet [96] which automatically fo854”
Object Detection in 20 Years: A Survey (2019, unknown): “To improve the detection of small objects, some “adaptive zoom-in” techniques are proposed in some recent detectors to adaptively enlarge the small objects into the “larger ones” [179], [180].”
Tree-Structured Reinforcement Learning for Sequential Object Localization (2016, background): “Perhaps [17] and [18] are the closest works to ours.”
Clustered Object Detection in Aerial Images (2019, methodology): “To handle this issue, we present the ScaleNet to estimate an appropriate scale for the objects in each cluster chip and then rescale the chip accordingly before feeding it to a detector, which is different from [10, 24, 18] by directly resizing cropped chips.”