Quantizing Oriented Object Detection Network via Outlier-Aware Quantization and IoU Approximation

Zhao, Mingxin; Ning, Ke; Yu, Shuangming; Liu, Liyuan; Wu, Nanjian

doi:10.1109/lsp.2020.3031490

Cited by 10 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 1 summarizes some classical post-training quantization schemes. Early PTQ focused on minimizing the quantization error of network parameters through techniques such as optimizing quantization factor scale [21,22], bias correction [27,28], piecewise linear quantization [29,30], and outlier separation [31,32]. For example, Nvidia's TensorRT [22], a widely used quantization tool, searched for the optimal quantization factor scale by minimizing the Kullback-Leibler (KL) distance between FP activation and quantized activation.…”

Section: Related Workmentioning

confidence: 99%

“…The Quantization Error of Network Parameters Optimizing Quantization Factor Scale [21,22] Bias Correction [27,28] Piecewise Linear Quantization [29,30] Outlier Separation [31,32] ≤4 bit Layer-wise Reconstruction LAPQ [23] AdaRound [24] AdaQuant [35] Block-wise Reconstruction BrecQ [25] RAPQ [33] Mr.BiQ [34] Qdrop [26] 3. Background and Theoretical Analysis 3.1.…”

Section: ≥6 Bitmentioning

confidence: 99%

See 1 more Smart Citation

AE-Qdrop: Towards Accurate and Efficient Low-Bit Post-Training Quantization for A Convolutional Neural Network

Li,

Chen,

Jin

et al. 2024

Electronics

View full text Add to dashboard Cite

Blockwise reconstruction with adaptive rounding helps achieve acceptable 4-bit post-training quantization accuracy. However, adaptive rounding is time intensive, and the optimization space of weight elements is constrained to a binary set, thus limiting the performance of quantized models. The optimality of block-wise reconstruction requires that subsequent network blocks remain unquantized. To address this, we propose a two-stage post-training quantization scheme, AE-Qdrop, encompassing block-wise reconstruction and global fine-tuning. In the block-wise reconstruction stage, a progressive optimization strategy is introduced as a replacement for adaptive rounding, enhancing both quantization accuracy and efficiency. Additionally, the integration of randomly weighted quantized activation helps mitigate the risk of overfitting. In the global fine-tuning stage, the weights of each quantized network block are corrected simultaneously through logit matching and feature matching. Experiments in image classification and object detection tasks validate that AE-Qdrop achieves high precision and efficient quantization. For the 2-bit MobileNetV2, AE-Qdrop outperforms Qdrop in quantization accuracy by 6.26%, and its quantization efficiency is fivefold higher.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: ≥6 Bitmentioning

confidence: 99%

AE-Qdrop: Towards Accurate and Efficient Low-Bit Post-Training Quantization for A Convolutional Neural Network

Li,

Chen,

Jin

et al. 2024

Electronics

View full text Add to dashboard Cite

show abstract

“…Both too many and too few anchors can lead to poor results, and excessive anchors can also increase computational complexity. Unfortunately, these algorithms use NMS during the detection process, rather than all edge devices supporting NMS (such as edge computing devices that only support integer operations) [37]. In order to solve the above problems and abandon manual intervention and the application of prior knowledge, researchers have begun to turn their attention to transformer-based DETR.…”

Section: Related Workmentioning

confidence: 99%

FCM-DETR: An Efficient End-to-End Fire Smoke and Human Detection based on Deformable DETR

Liang,

Zeng

2024

Preprint

View full text Add to dashboard Cite

Fire is a serious security threat that can lead to casualties, property damage, and environmental damage. However, despite the availability of object detection algorithms, challenges persist in detecting fires and smoke. These challenges include slow convergence speed, poor performance in detecting small targets, and high computational cost limiting deployments. In this paper, Fire smoke and human detection based on ConvNeXt and Mixed encoder (FCM-DETR), an end-to-end object detection algorithm based on Deformable DETR, is proposed. Firstly, we introduce ConvNeXt to take the place of Resnet, which greatly reduces the amount of computation and improves the ability to extract irregular flame features. Secondly, to effectively process multi-scale features, the original encoder is decoupled into two modules. Then Mixed encoder, an innovative encoder structure, is proposed, resulting in excellent performance of multi-scale fire and smoke features fusion. What can’t be overlooked is that the encoder block is compatible with any DETR-based models. Finally, the convergence speed is accelerated and the mean of average precision is improved by applying a novel loss function called Powerful IoU v2. The experimental results indicate that our model achieves the best detection accuracy compared to other models, with mAP reaching 66.7%, mAPs and Accuracyfire achieving impressive 50.2% and 98.05%, respectively.

show abstract

“…Both too many and too few anchors lead to poor results, and excessive anchors also increase computational complexity. Unfortunately, these algorithms use NMS during the detection process, rather than all edge devices supporting NMS (such as edge computing devices that only support integer operations) [44]. In order to solve the above problems and abandon manual intervention and the application of prior knowledge, researchers have begun to turn their attention to transformer-based DETR.…”

Section: Related Workmentioning

confidence: 99%

FSH-DETR: An Efficient End-to-End Fire Smoke and Human Detection Based on a Deformable DEtection TRansformer (DETR)

Liang,

Zeng

2024

Sensors

View full text Add to dashboard Cite

Fire is a significant security threat that can lead to casualties, property damage, and environmental damage. Despite the availability of object-detection algorithms, challenges persist in detecting fires, smoke, and humans. These challenges include poor performance in detecting small fires and smoke, as well as a high computational cost, which limits deployments. In this paper, we propose an end-to-end object detector for fire, smoke, and human detection based on Deformable DETR (DEtection TRansformer) called FSH-DETR. To effectively process multi-scale fire and smoke features, we propose a novel Mixed Encoder, which integrates SSFI (Separate Single-scale Feature Interaction Module) and CCFM (CNN-based Cross-scale Feature Fusion Module) for multi-scale fire, smoke, and human feature fusion. Furthermore, we enhance the convergence speed of FSH-DETR by incorporating a bounding box loss function called PIoUv2 (Powerful Intersection of Union), which improves the precision of fire, smoke, and human detection. Extensive experiments on the public dataset demonstrate that the proposed method surpasses state-of-the-art methods in terms of the mAP (mean Average Precision), with mAP and mAP50 reaching 66.7% and 84.2%, respectively.

show abstract

Quantizing Oriented Object Detection Network via Outlier-Aware Quantization and IoU Approximation

Cited by 10 publications

References 8 publications

AE-Qdrop: Towards Accurate and Efficient Low-Bit Post-Training Quantization for A Convolutional Neural Network

AE-Qdrop: Towards Accurate and Efficient Low-Bit Post-Training Quantization for A Convolutional Neural Network

FCM-DETR: An Efficient End-to-End Fire Smoke and Human Detection based on Deformable DETR

FSH-DETR: An Efficient End-to-End Fire Smoke and Human Detection Based on a Deformable DEtection TRansformer (DETR)

Contact Info

Product

Resources

About