NAS-FCOS: Fast Neural Architecture Search for Object Detection

Wang, Ning; Gao, Yang; Chen, Hao; Wang, Peng; Tian, Zhi; Shen, Chunhua; Zhang, Yanning

doi:10.48550/arxiv.1906.04423

Cited by 15 publications

(22 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Even the automatic learning based method, a.k.a. NAS-FCOS [36] that performs 1.6 lower than the baseline. This phenomenon highlights the domain gap between general object and face detection.…”

Section: Introductionmentioning

confidence: 93%

“…4.1.1 AutoFA. In order to address the limitation of previous NASbased FPN [10,36,37] when applied on the face domain, the above analysis motivates us to design a module that aggregates a feature by the similar-scale features instead of directly using those with large differences in scale.…”

Section: Search Space Of Autofa and Autofementioning

confidence: 99%

“…For another instance, ASPP [2,25], RFB [20] and RFE [8] modules are proposed to enhance the feature representation by adjusting the effective receptive fields. Recently, Neural Architecture Search (NAS) has been also investigated for object detection, which has achieved remarkable performance gains, such as NAS-FPN [10], AutoFPN [37] and NAS-FCOS [36]. However, such a gain is severely not generalized when applying to face detection.…”

Section: Introductionmentioning

confidence: 99%

“…[25] proposes a novel recursive FPN that incorporates extra feedback connections from FPN into the bottomup backbone layers. Nowadays, NAS-based methods have demonstrated much success in exploring a better architecture for feature fusion and refinement [10,36,37]. Besides, feature enhancement modules are also be widely studied.…”

Section: Introductionmentioning

confidence: 99%

“…As for their applications on object detection to fuse the multi-scale features, NAS-FPN [10] searches the irregular connections among pyramid layers with an RNN controller for aggregating the multi-scale features. AutoFPN [37] and NAS-FCOS [36] discover the aggregation modules within a fully-connected search space densely connecting any two layers, in which features from some layers that damage the aggregated feature may be introduced causing accuracy degradation. BFBox [23] is the first attempt of NAS on face detection and proposes a face-suitable search space.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

ASFD: Automatic and Scalable Face Detector

Zhang

Wang

et al. 2021

Proceedings of the 29th ACM International Conference on Multimedia

View full text Add to dashboard Cite

Along with current multi-scale based detectors, Feature Aggregation and Enhancement (FAE) modules have shown superior performance gains for cutting-edge object detection. However, these hand-crafted FAE modules show inconsistent improvements on face detection, which is mainly due to the significant distribution difference between its training and applying corpus, i.e. COCO vs. WIDER Face. To tackle this problem, we essentially analyse the effect of data distribution, and consequently propose to search an effective FAE architecture, termed AutoFAE by a differentiable architecture search, which outperforms all existing FAE modules in face detection with a considerable margin. Upon the found Auto-FAE and existing backbones, a supernet is further built and trained, which automatically obtains a family of detectors under the different complexity constraints. Extensive experiments conducted on popular benchmarks, i.e. WIDER Face and FDDB, demonstrate the state-of-the-art performance-efficiency trade-off for the proposed automatic and scalable face detector (ASFD) family. In particular, our strong ASFD-D6 outperforms the best competitor with AP 96.7/96.2/92.1 on WIDER Face test, and the lightweight ASFD-D0 costs about 3.1 ms, i.e. more than 320 FPS, on the V100 GPU with VGA-resolution images. CCS CONCEPTS• Computing methodologies → Object detection.

show abstract

Section: Introductionmentioning

confidence: 93%

Section: Search Space Of Autofa and Autofementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

ASFD: Automatic and Scalable Face Detector

Zhang

Wang

et al. 2021

Proceedings of the 29th ACM International Conference on Multimedia

View full text Add to dashboard Cite

show abstract

NASTransfer: Analyzing Architecture Transferability in Large Scale Neural Architecture Search

Panda

Merler²,

Jaiswal³

et al. 2021

AAAI

View full text Add to dashboard Cite

Neural Architecture Search (NAS) is an open and challenging problem in machine learning. While NAS offers great promise, the prohibitive computational demand of most of the existing NAS methods makes it difficult to directly search the architectures on large-scale tasks. The typical way of conducting large scale NAS is to search for an architectural building block on a small dataset (either using a proxy set from the large dataset or a completely different small scale dataset) and then transfer the block to a larger dataset. Despite a number of recent results that show the promise of transfer from proxy datasets, a comprehensive evaluation of different NAS methods studying the impact of different source datasets has not yet been addressed. In this work, we propose to analyze the architecture transferability of different NAS methods by performing a series of experiments on large scale benchmarks such as ImageNet1K and ImageNet22K. We find that: (i) The size and domain of the proxy set does not seem to influence architecture performance on the target dataset. On average, transfer performance of architectures searched using completely different small datasets (e.g., CIFAR10) perform similarly to the architectures searched directly on proxy target datasets. However, design of proxy sets has considerable impact on rankings of different NAS methods. (ii) While different NAS methods show similar performance on a source dataset (e.g., CIFAR10), they significantly differ on the transfer performance to a large dataset (e.g., ImageNet1K). (iii) Even on large datasets, random sampling baseline is very competitive, but the choice of the appropriate combination of proxy set and search strategy can provide significant improvement over it. We believe that our extensive empirical analysis will prove useful for future design of NAS algorithms.

show abstract

SM-NAS: Structural-to-Modular Neural Architecture Search for Object Detection

Yao

Zhang

et al. 2019

Preprint

View full text Add to dashboard Cite

The state-of-the-art object detection method is complicated with various modules such as backbone, feature fusion neck, RPN, and RCNN head, where each module may have different designs and structures. How to leverage the computational cost and accuracy trade-off for the structural combination as well as the modular selection of multiple modules? Neural architecture search (NAS) has shown great potential in finding an optimal solution. Existing NAS works for object detection only focus on searching better design of a single module such as backbone or feature fusion neck, while neglecting the balance of the whole system. In this paper, we present a two-stage coarse-to-fine searching strategy named Structural-to-Modular NAS (SM-NAS) for searching a GPU-friendly design of both an efficient combination of modules and better modular-level architecture for object detection. Specifically, Structural-level searching stage first aims to find an efficient combination of different modules; Modular-level searching stage then evolves each specific module and pushes the Pareto front forward to a faster task-specific network. We consider a multi-objective search where the search space covers many popular designs of detection methods. We directly search a detection backbone without pre-trained models or any proxy task by exploring a fast training from scratch strategy. The resulting architectures dominate state-of-the-art object detection systems in both inference time and accuracy and demonstrate the effectiveness on multiple detection datasets, e.g. halving the inference time with additional 1% mAP improvement compared to FPN and reaching 46% mAP with the similar inference time of MaskRCNN.

show abstract

NAS-FCOS: Fast Neural Architecture Search for Object Detection

Cited by 15 publications

References 25 publications

ASFD: Automatic and Scalable Face Detector

ASFD: Automatic and Scalable Face Detector

NASTransfer: Analyzing Architecture Transferability in Large Scale Neural Architecture Search

SM-NAS: Structural-to-Modular Neural Architecture Search for Object Detection

Contact Info

Product

Resources

About