Efficient Object Detection in Large Images Using Deep Reinforcement Learning

Uzkent, Burak; Yeh, Christopher; Ermon, Stefano

doi:10.1109/wacv45572.2020.9093447

Cited by 94 publications

(51 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For comparison, we explore using four representative object detection models with a proposed model for the HRSID dataset, which are Faster R-CNN [16], RetinaNet [17], YOLOv3 [9], YOLOv5 [18], and HRSDNet [11]. And Ef-ficientOD [8] is considered as baseline. EfficientOD and YOLO baselines, we follow parameters from publicly available code except for the size of input images.…”

Section: Methodsmentioning

confidence: 99%

“…Previous methods proposed to overcome these limitations fall into one of the following two categories: 1) utilization of the lightweight model architectures [3,4,5], and 2) acceleration of the operation time by reducing the search space itself [6,7]. On the other hand, Uzkent et al [8] proposed an efficient detection model to use LR images as much as possible, as long as sufficient information can be incorporated by a Reinforcement Learning (RL) agent and YOLOv3 [9] detectors. However, this approach cannot be applied for an end-to-end framework and requires pre-trained detection models.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Sarod: Efficient End-To-End Object Detection On SAR Images With Reinforcement Learning

Kang

Jeon

Bang

et al. 2021

2021 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

Generally, object detection on Synthetic-Aperture Radar (SAR) images is known to be more challenging than that in Electro-Optical (EO) satellite images because SAR images have non-negligible speckle noise and require extensive data pre-processing. Nevertheless, object detection in SAR images is important, as SAR imagery can be obtained under severe weather and time conditions. While many recent object detection approaches on SAR imagery focus on improving detection accuracy, few studies focus on improving processing efficiency. In fact, there are significant challenges and trade-offs to achieve both high accuracy and efficiency at the same time. In this work, we introduce SAROD, a novel efficient end-to-end object detection framework on SAR images based on Reinforcement Learning (RL) to balance the tradeoffs. Our proposed model consists of two detectors, coarse and fine-grained detectors, with an RL agent, where RL has not yet been utilized for object detection on SAR imagery. Our model was evaluated on a challenging SAR imagery dataset, achieving performance comparable to state-of-the-art detectors while maintaining high efficiency of source data usage.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Sarod: Efficient End-To-End Object Detection On SAR Images With Reinforcement Learning

Kang

Jeon

Bang

et al. 2021

2021 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

show abstract

“…Reinforcement learning based Region Proposal Network (RPN) [10] has also been explored to increase its efficiency of computation. In [79], [80], an agent selects the regions of the image which need to be processed at higher resolution or by a finer detector to reduce computation. In [81] RoIs are effectively selected based on the interdependence of objects with tree-structured reinforcement learning.…”

Section: Single-image Roi Selection Methodsmentioning

confidence: 99%

Adaptive Inattentional Framework for Video Object Detection With Reward-Conditional Training

et al. 2020

View full text Add to dashboard Cite

Recent object detection studies have been focused on video sequences, mostly due to the increasing demand of industrial applications. Although single-image architectures achieve remarkable results in terms of accuracy, they do not take advantage of particular properties of the video sequences and usually require high parallel computational resources, such as desktop GPUs. In this work, an inattentional framework is proposed, where the object context in video frames is dynamically reused in order to reduce the computation overhead. The context features corresponding to keyframes are fused into a synthetic feature map, which is further refined using temporal aggregation with ConvLSTMs. Furthermore, an inattentional policy has been learned to adaptively balance the accuracy and the amount of context reused. The inattentional policy has been learned under the reinforcement learning paradigm, and using our novel reward-conditional training scheme, which allows for policy training over a whole distribution of reward functions and enables the selection of a unique reward function at inference time. Our framework shows outstanding results on platforms with reduced parallelization capabilities, such as CPUs, achieving an average latency reduction up to 2.09x, and obtaining FPS rates similar to their equivalent GPU platform, at the cost of a 1.11x mAP reduction.

show abstract

“…For example, Ref. [89] trained an agent to select appropriate spatial resolutions for satellite images that run through detectors. When a low spatial resolution image is dominated by large objects, the agent passes this image to a coarse level detector.…”

Section: Spatial Samplingmentioning

confidence: 99%

Machine Learning of Spatial Data

Nikparvar

Thill

2021

IJGI

View full text Add to dashboard Cite

Properties of spatially explicit data are often ignored or inadequately handled in machine learning for spatial domains of application. At the same time, resources that would identify these properties and investigate their influence and methods to handle them in machine learning applications are lagging behind. In this survey of the literature, we seek to identify and discuss spatial properties of data that influence the performance of machine learning. We review some of the best practices in handling such properties in spatial domains and discuss their advantages and disadvantages. We recognize two broad strands in this literature. In the first, the properties of spatial data are developed in the spatial observation matrix without amending the substance of the learning algorithm; in the other, spatial data properties are handled in the learning algorithm itself. While the latter have been far less explored, we argue that they offer the most promising prospects for the future of spatial machine learning.

show abstract

Efficient Object Detection in Large Images Using Deep Reinforcement Learning

Cited by 94 publications

References 34 publications

Sarod: Efficient End-To-End Object Detection On SAR Images With Reinforcement Learning

Sarod: Efficient End-To-End Object Detection On SAR Images With Reinforcement Learning

Adaptive Inattentional Framework for Video Object Detection With Reward-Conditional Training

Machine Learning of Spatial Data

Contact Info

Product

Resources

About