Open AccessProceedings Article
Loss Function Discovery for Object Detection via Convergence-Simulation Driven Search
Reads0
Chats0
TLDR
CSE-Autoloss as discussed by the authors proposes an effective convergence-simulation driven evolutionary search algorithm, which can accelerate the search progress by regularizing the mathematical rationality of loss candidates via two progressive convergence simulation modules: convergence property verification and model optimization simulation.Abstract:
Designing proper loss functions for vision tasks has been a long-standing research direction to advance the capability of existing models For object detection, the well-established classification and regression loss functions have been carefully designed by considering diverse learning challenges (eg class imbalance, hard negative samples, and scale variances) Inspired by the recent progress in network architecture search, it is interesting to explore the possibility of discovering new loss function formulations via directly searching the primitive operation combinations So that the learned losses not only fit for diverse object detection challenges to alleviate huge human efforts, but also have better alignment with evaluation metric and good mathematical convergence property Beyond the previous auto-loss works on face recognition and image classification, our work makes the first attempt to discover new loss functions for the challenging object detection from primitive operation levels and finds the searched losses are insightful We propose an effective convergence-simulation driven evolutionary search algorithm, called CSE-Autoloss, for speeding up the search progress by regularizing the mathematical rationality of loss candidates via two progressive convergence simulation modules: convergence property verification and model optimization simulation CSE-Autoloss involves the search space (ie 21 mathematical operators, 3 constant-type inputs, and 3 variable-type inputs) that cover a wide range of the possible variants of existing losses and discovers best-searched loss function combination within a short time (around 15 wall-clock days with 20x speedup in comparison to the vanilla evolutionary algorithm) We conduct extensive evaluations of loss function search on popular detectors and validate the good generalization capability of searched losses across diverse architectures and various datasets Our experiments show that the best-discovered loss function combinations outperform default combinations (Cross-entropy/Focal loss for classification and L1 loss for regression) by 11% and 08% in terms of mAP for two-stage and one-stage detectors on COCO respectively Our searched losses are available at https://githubcom/PerdonLiu/CSE-Autolossread more
Citations
More filters
Posted Content
AutoLoss-Zero: Searching Loss Functions from Scratch for Generic Tasks
TL;DR: In this article, the authors proposed AutoLoss-Zero, a general framework for searching loss functions from scratch for generic tasks, which can generalize well to different datasets and networks.
Posted Content
Rank & Sort Loss for Object Detection and Instance Segmentation
TL;DR: Rank & Sort (RS) Loss as discussed by the authors is a ranking-based loss function to train deep object detection and instance segmentation methods (i.e. visual detectors), which supervises the classifier, a sub-network of these methods, to rank each positive above all negatives as well as to sort positives among themselves with respect to their continuous localisation qualities (e.g. intersection-over-union - IoU).
Posted Content
Problem Learning: Towards the Free Will of Machines.
TL;DR: In this article, the authors formalize problem learning as the identification of valid and ethical problems in a problem space and introduce several possible approaches to problem learning, and discuss the ethical implications of problem learning under the background of Responsible AI.
Posted Content
HLIC: Harmonizing Optimization Metrics in Learned Image Compression by Reinforcement Learning
TL;DR: In this paper, the authors propose to harmonize optimization metrics in HLIC using online loss function adaptation by reinforcement learning, achieving better visual quality and higher VMAF score by leveraging the advantages of both peak signal-to-noise ratio and MS-SSIM.
Posted Content
Tightening the Approximation Error of Adversarial Risk with Auto Loss Function Search.
Pengfei Xia,Ziqiang Li,Bin Li +2 more
TL;DR: In this article, the authors proposed AutoLoss-AR, a method for searching loss functions for tightening the approximation error of adversarial risk, which outperformed the handcrafted baseline by 0.9-2.9% and 0.7-2% on MNIST and CIFAR-10.
References
More filters
Proceedings ArticleDOI
Deep Residual Learning for Image Recognition
TL;DR: In this article, the authors proposed a residual learning framework to ease the training of networks that are substantially deeper than those used previously, which won the 1st place on the ILSVRC 2015 classification task.
Book ChapterDOI
Microsoft COCO: Common Objects in Context
Tsung-Yi Lin,Michael Maire,Serge Belongie,James Hays,Pietro Perona,Deva Ramanan,Piotr Dollár,C. Lawrence Zitnick +7 more
TL;DR: A new dataset with the goal of advancing the state-of-the-art in object recognition by placing the question of object recognition in the context of the broader question of scene understanding by gathering images of complex everyday scenes containing common objects in their natural context.
Proceedings ArticleDOI
Feature Pyramid Networks for Object Detection
TL;DR: This paper exploits the inherent multi-scale, pyramidal hierarchy of deep convolutional networks to construct feature pyramids with marginal extra cost and achieves state-of-the-art single-model results on the COCO detection benchmark without bells and whistles.
Proceedings ArticleDOI
Fast R-CNN
TL;DR: Fast R-CNN as discussed by the authors proposes a Fast Region-based Convolutional Network method for object detection, which employs several innovations to improve training and testing speed while also increasing detection accuracy and achieves a higher mAP on PASCAL VOC 2012.
Proceedings Article
Faster R-CNN: towards real-time object detection with region proposal networks
TL;DR: Ren et al. as discussed by the authors proposed a region proposal network (RPN) that shares full-image convolutional features with the detection network, thus enabling nearly cost-free region proposals.