Pith. sign in

REVIEW 6 cited by

Optimizing YOLOv5s Object Detection through Knowledge Distillation algorithm

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2410.12259 v1 pith:7YDIJCVV submitted 2024-10-16 cs.CV cs.LG

Guanming Huang , Aoran Shen , Yuxiang Hu , Junliang Du , Jiacheng Hu , Yingbin Liang This is my paper · ORCID
classification cs.CVcs.LG
keywords distillationmodeldetectionaccuracyknowledgestudentyolov5scurve
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

This paper explores the application of knowledge distillation technology in target detection tasks, especially the impact of different distillation temperatures on the performance of student models. By using YOLOv5l as the teacher network and a smaller YOLOv5s as the student network, we found that with the increase of distillation temperature, the student's detection accuracy gradually improved, and finally achieved mAP50 and mAP50-95 indicators that were better than the original YOLOv5s model at a specific temperature. Experimental results show that appropriate knowledge distillation strategies can not only improve the accuracy of the model but also help improve the reliability and stability of the model in practical applications. This paper also records in detail the accuracy curve and loss function descent curve during the model training process and shows that the model converges to a stable state after 150 training cycles. These findings provide a theoretical basis and technical reference for further optimizing target detection algorithms.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Self-Supervised Learning in Deep Networks: A Pathway to Robust Few-Shot Classification

    cs.CV 2024-11 reject novelty 3.0 of 10

    A report claiming 95.12% few-shot accuracy on Mini-ImageNet from a self-supervised ResNet-101 pipeline, with insufficient experimental evidence.

  2. Leveraging Generative Adversarial Networks for Addressing Data Imbalance in Financial Market Supervision

    q-fin.CP 2024-12 reject novelty 2.0 of 10

    A standard GAN is used to balance a financial dataset, and the paper reports small accuracy improvements over traditional sampling methods, though without sufficient experimental support.

  3. Optimizing Gesture Recognition for Seamless UI Interaction Using Convolutional Neural Networks

    cs.HC 2024-11 reject novelty 2.0 of 10

    A routine CNN benchmark for 14 hand gestures reports AUC 0.83 and recall 0.85 for an undescribed Ours model, with no error bars or code.

  4. Deep Learning for Cross-Border Transaction Anomaly Detection in Anti-Money Laundering Systems

    cs.LG 2024-11 reject novelty 2.0 of 10

    A proposed CRNIM hybrid CNN-GRU model is claimed to reach 97.1% accuracy and 0.94 AUROC on the Elliptic Bitcoin dataset for anomaly detection.

  5. Adaptive Cache Management for Complex Storage Systems Using CNN-LSTM-Based Spatiotemporal Prediction

    cs.DC 2024-11 reject novelty 2.0 of 10

    A CNN-LSTM model is claimed to predict storage cache demand better than six baselines, but the only numerical evidence is a single table without validation details.

  6. A Combined Encoder and Transformer Approach for Coherent and High-Quality Text Generation

    cs.CL 2024-11 reject novelty 2.0 of 10

    A proposed BERT-plus-GPT-4 hybrid is claimed to beat GPT-3, T5, BART, Transformer-XL, and CTRL on perplexity and BLEU, but the experiments are not reproducible.

Pith tools