高级搜索

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

文本语义引导的动态课程式学习类增量目标检测方法

高鹏,  张若楠,  刘立波

高鹏, 张若楠, 刘立波. 文本语义引导的动态课程式学习类增量目标检测方法[J]. 电子与信息学报. doi: 10.11999/JEIT260987
引用本文: 高鹏, 张若楠, 刘立波. 文本语义引导的动态课程式学习类增量目标检测方法[J]. 电子与信息学报. doi: 10.11999/JEIT260987
GAO Peng, ZHANG Ruonan, LIU Libo. Text Semantic-Guided Dynamic Curriculum Learning for Class-Incremental Object Detection[J]. Journal of Electronics & Information Technology. doi: 10.11999/JEIT260987
Citation: GAO Peng, ZHANG Ruonan, LIU Libo. Text Semantic-Guided Dynamic Curriculum Learning for Class-Incremental Object Detection[J]. Journal of Electronics & Information Technology. doi: 10.11999/JEIT260987

文本语义引导的动态课程式学习类增量目标检测方法

doi: 10.11999/JEIT260987 cstr: 32379.14.JEIT260987
基金项目: 宁夏自然科学基金(No. 2024AAC02010),宁夏科技创新领军人才项目(No. 2022GKLRLX03),国家自然科学基金(No. 62262053, No. 62506179),银川市科技项目(No. 2025RC09),宁夏大学研究生创新项目(CXXM2026127)
详细信息
    作者简介:

    高鹏:男,硕士生,研究方向为目标检测、增量学习等

    张若楠:女,博士,副教授,研究领域为人机感知共友好的多媒体处理与分析

    刘立波:女,博士,教授,博士生导师,研究方向为智能信息处理、计算机视觉等

    通讯作者:

    刘立波 liulib@nxu.edu.cn

  • 中图分类号: TP391.41

Text Semantic-Guided Dynamic Curriculum Learning for Class-Incremental Object Detection

Funds: Natural Science Foundation of Ningxia, China (No. 2024AAC02010), Science and Technology Leading Talent Training Projects of Ningxia (No. 2022GKLRLX03), National Natural Science Foundation of China (No. 62262053, No. 62506179), Yinchuan City Science and Technology Projects (No. 2025RC09), Postgraduate Innovation Project of Ningxia University (CXXM2026127)
  • 摘要: 类增量目标检测旨在持续学习新类别的同时,保持对已学习旧类别的检测能力。现有方法存在目标类别判别能力不足、旧类伪标签可靠性低以及新旧知识难以平衡等问题。为此,提出一种文本语义引导的动态课程式学习类增量目标检测方法。首先构建基于语义先验的判别模块,利用类别文本原型与视觉语义特征的相似度完成分类,增强模型类别判别能力;其次提出视觉文本辅助的伪标签筛选模块,通过视觉语言问答判断教师预测框的旧类归属性,仅保留语义一致且无冲突的旧类预测参与回放监督,减少低质量伪标签干扰;最后设计动态自适应课程式蒸馏策略,根据新类检测损失变化自适应触发旧类蒸馏,并逐步提升蒸馏权重,实现新类学习与旧类知识保持的平滑过渡。在单阶段增量场景中,所提方法在COCO数据集70+10和40+40的划分下,AP较基线分别提升1.8%和2.1%;在HiFSOD-Bird数据集1074+358和716+716的划分下,AP较基线分别提升2.4%和2.2%,并且该方法在多阶段增量场景中也优于多种主流方法。实验结果表明,所提方法能够有效缓解灾难性遗忘,在粗粒度和细粒度增量场景中均具有较好的检测精度、稳定性和鲁棒性。
  • 图  1  TD-DETR模型框架图

    图  2  语义先验的判别模块

    图  3  视觉文本辅助的伪标签筛选模块

    图  4  HiFSOD-Bird数据集拼接过程

    图  5  检测结果可视化对比

    注:紫色背景为粗粒度,橙色背景为细粒度;绿色框为基线误检目标,红色框为模型的检测框

    图  6  伪标签生成可视化对比

    注:紫色背景为粗粒度,橙色背景为细粒度;蓝色框为基线生成的错误伪标签,红色框为模型生成的伪标签框

    表  1  COCO数据集对比实验结果(%)

    设置方法检测器APAP50AP75APSAPMAPL
    70+10RILOD19GFLv124.537.925.714.227.433.5
    SID21GFLv132.849.035.017.136.944.5
    ERD23GFLv134.951.937.418.738.845.5
    IOR25GFLv135.649.336.621.438.942.4
    CL-DETR23Deformable DETR39.056.842.422.642.251.7
    DiffKA25Deformable DETR40.357.644.122.944.053.3
    CC-DETR26Deformable DETR40.958.844.323.943.753.9
    DyQ-DETR25Deformable DETR41.960.145.824.145.355.8
    IAQD25Deformable DETR42.260.546.225.145.255.6
    BPF24Faster R-CNN36.256.838.9---
    RGR25Faster R-CNN36.656.639.6---
    Our(TD-DETR)Deformable DETR42.760.645.626.645.256.4
    40+40RILOD19GFLv129.945.032.015.833.040.5
    SID21GFLv134.051.436.318.438.444.9
    ERD23GFLv136.954.539.621.340.447.5
    IOR25GFLv135.149.937.420.837.844.6
    CL-DETR23Deformable DETR36.754.339.020.239.850.1
    DiffKA25Deformable DETR37.255.139.819.440.050.3
    CC-DETR26Deformable DETR39.156.642.121.640.753.8
    DyQ-DETR25Deformable DETR39.757.543.021.642.953.8
    IAQD25Deformable DETR39.957.743.422.943.052.2
    BPF24Faster R-CNN34.454.337.3---
    RGR25Faster R-CNN35.656.038.2---
    Our(TD-DETR)Deformable DETR41.259.144.624.544.354.0
    注:方法上角标为所提模型发表年限,如CL-DETR23为2023年发表
    下载: 导出CSV

    表  2  COCO数据集多阶段对比实验结果(%)

    方法 40+20+20 40+10+10+10+10
    1 2 1 2 3 4
    SID21 34.0 23.8 34.6 24.1 14.6 12.6
    BPF24 34.2 23.4 34.7 24.1 14.6 12.7
    IOR25 36.1 32.5 36.4 30.8 26.2 20.7
    ERD23 36.7 32.4 36.4 30.8 26.2 20.7
    DiffKA25 36.8 36.1 36.6 32.7 31.6 29.6
    CL-DETR23 37.2 35.3 36.5 33.4 30.8 28.1
    CC-DETR26 40.7 38.2 41.3 38.7 35.4 33.6
    Our(TD-DETR) 43.4 40.1 42.8 39.3 36.7 34.8
    下载: 导出CSV

    表  3  HiFSOD-Bird数据集对比实验结果(%)

    设置方法检测器APAP50AP75ARAR50AR75
    1074+358IOR25GFLv138.748.845.664.664.964.9
    BPF24Faster R-CNN39.449.846.665.265.765.7
    CL-DETR23Deformable DETR41.852.949.871.573.273.3
    CC-DETR26Deformable DETR44.155.252.171.673.373.3
    Our(TD-DETR)Deformable DETR46.559.455.972.974.574.5
    716+716IOR25GFLv138.449.045.864.565.165.1
    BPF24Faster R-CNN40.951.548.066.667.167.1
    CL-DETR23Deformable DETR43.054.751.268.970.370.3
    CC-DETR26Deformable DETR45.057.153.671.072.872.8
    Our(TD-DETR)Deformable DETR47.260.356.473.675.575.5
    下载: 导出CSV

    表  4  HiFSOD-Bird数据集多阶段对比实验结果(%)

    方法716+358+358716+179+179+179+179
    121234
    BPF2437.731.238.131.425.319.8
    IOR2537.230.638.131.425.319.8
    CL-DETR2338.931.438.332.026.721.6
    CC-DETR2641.533.141.433.728.624.4
    Our(TD-DETR)43.836.143.537.132.429.1
    下载: 导出CSV

    表  5  COCO蒸馏介入早晚实验对比(%)

    状态λemaxAPAPoldAPnew
    介入早0.71542.344.531.5
    介入晚0.32542.044.132.8
    介入适中0.52042.744.432.2
    下载: 导出CSV

    表  6  COCO数据集蒸馏权重增加缓急实验对比(%)

    状态rωmaxAPAPoldAPnew
    权重增长较慢12342.243.431.7
    权重增长较快8542.443.430.6
    权重增长适中10442.744.432.2
    下载: 导出CSV

    表  7  COCO数据集IOU阈值实验对比(%)

    δAPAP50AP75APSAPMAPL
    0.641.859.945.025.844.955.1
    0.742.760.645.626.645.256.4
    0.842.360.145.226.245.055.8
    下载: 导出CSV

    表  10  COCO数据集文本模板不同数量对比(%)

    模板数编码文本总数生成时间/sAPAPoldAPnew
    1920.2141.943.031.7
    32760.2842.744.432.2
    54600.3542.844.532.0
    下载: 导出CSV

    表  8  COCO数据集消融实验结果(%)

    实验序号方法SPDMVTPFDACDAPAPoldAPnew
    1Baseline(CC-DETR)40.942.531.0
    2+SPDM√41.442.731.8
    3+VTPF√√42.143.931.9
    4Ours(TD-DETR)√√√42.744.432.2
    下载: 导出CSV

    表  9  HiFSOD-Bird数据集消融实验结果(%)

    实验序号方法SPDMVTPFDACDAPAPoldAPnew
    1Baseline(CC-DETR)44.143.352.3
    2+SPDM√44.943.854.1
    3+VTPF√√46.044.754.8
    4Ours(TD-DETR)√√√46.545.255.6
    下载: 导出CSV
  • [1] 梁燕, 杨会林, 邵凯. 自适应特征选择的车路协同3D目标检测方案[J]. 电子与信息学报, 2025, 47(12): 5214–5225. doi: 10.11999/JEIT250601.

    LIANG Yan, YANG Huilin, and SHAO Kai. A vehicle-infrastructure cooperative 3D object detection scheme based on adaptive feature selection[J]. Journal of Electronics & Information Technology, 2025, 47(12): 5214–5225. doi: 10.11999/JEIT250601.
    [2] 吕晓华, 魏铭辰, 刘立波. 基于位置可学习视觉中心机制的零售商品检测方法[J]. 物联网学报, 2023, 7(4): 142–152. doi: 10.11959/j.issn.2096-3750.2023.00366.

    LYU Xiaohua, WEI Mingchen, and LIU Libo. Retail commodity detection method based on location learnable visual center mechanism[J]. Chinese Journal on Internet of Things, 2023, 7(4): 142–152. doi: 10.11959/j.issn.2096-3750.2023.00366.
    [3] 翁星星, 庞超, 许博文, 等. 面向遥感图像解译的增量深度学习[J]. 电子与信息学报, 2024, 46(10): 3979–4001. doi: 10.11999/JEIT240172.

    WENG Xingxing, PANG Chao, XU Bowen, et al. Incremental deep learning for remote sensing image interpretation[J]. Journal of Electronics & Information Technology, 2024, 46(10): 3979–4001. doi: 10.11999/JEIT240172.
    [4] 李斌, 崔宗勇, 汪浩瀚, 等. SAR目标增量识别中基于最大化非重合体积的样例挑选方法[J]. 电子与信息学报, 2024, 46(10): 3918–3927. doi: 10.11999/JEIT240217.

    LI Bin, CUI Zongyong, WANG Haohan, et al. Exemplar selection based on maximizing non-overlapping volume in SAR target incremental recognition[J]. Journal of Electronics & Information Technology, 2024, 46(10): 3918–3927. doi: 10.11999/JEIT240217.
    [5] 张东阳, 陆子轩, 刘军民, 等. 深度模型的持续学习综述: 理论、方法和应用[J]. 电子与信息学报, 2024, 46(10): 3849–3878. doi: 10.11999/JEIT240095.

    ZHANG Dongyang, LU Zixuan, LIU Junmin, et al. A survey of continual learning with deep networks: Theory, method and application[J]. Journal of Electronics & Information Technology, 2024, 46(10): 3849–3878. doi: 10.11999/JEIT240095.
    [6] SHMELKOV K, SCHMID C, and ALAHARI K. Incremental learning of object detectors without catastrophic forgetting[C]. Proceedings of 2017 IEEE International Conference on Computer Vision, Venice, Italy, 2017: 3420–3429. doi: 10.1109/ICCV.2017.368.
    [7] PENG Can, ZHAO Kun, and LOVELL B C. Faster ILOD: Incremental learning for object detectors based on faster RCNN[J]. Pattern Recognition Letters, 2020, 140: 109–115. doi: 10.1016/j.patrec.2020.09.030.
    [8] FENG Tao, WANG Mang, and YUAN Hangjie. Overcoming catastrophic forgetting in incremental object detection via elastic response distillation[C]. Proceedings of 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, USA, 2022: 9427–9436. doi: 10.1109/CVPR52688.2022.00921.
    [9] PENG Can, ZHAO Kun, MAKSOUD S, et al. DIODE: Dilatable incremental object detection[J]. Pattern Recognition, 2023, 136: 109244. doi: 10.1016/j.patcog.2022.109244.
    [10] ZHU Xizhou, SU Weijie, LU Lewei, et al. Deformable DETR: Deformable transformers for end-to-end object detection[C]. Proceedings of the 9th International Conference on Learning Representations, 2021. (查阅网上资料, 未找到对应的出版地信息, 请确认).
    [11] LIU Yaoyao, SCHIELE B, VEDALDI A, et al. Continual detection transformer for incremental object detection[C]. Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, Canada, 2023: 23799–23808. doi: 10.1109/CVPR52729.2023.02279.
    [12] ZHANG Jichuan, LI Wei, CHENG Shuang, et al. Dynamic object queries for transformer-based incremental object detection[C]. Proceedings of 2025 IEEE International Conference on Acoustics, Speech and Signal Processing, Hyderabad, India, 2025: 1–5. doi: 10.1109/ICASSP49660.2025.10888317.
    [13] RADFORD A, KIM J W, HALLACY C, et al. Learning transferable visual models from natural language supervision[C]. Proceedings of the 38th International Conference on Machine Learning, 2021: 8748–8763. (查阅网上资料, 未找到对应的出版地信息, 请确认).
    [14] LI Dawei, TASCI S, GHOSH S, et al. RILOD: Near real-time incremental learning for object detection at the edge[C]. Proceedings of the 4th ACM/IEEE Symposium on Edge Computing, Arlington, USA, 2019: 113–126. doi: 10.1145/3318216.3363317.
    [15] PENG Can, ZHAO Kun, MAKSOUD S, et al. SID: Incremental learning for anchor-free object detection via selective and inter-related distillation[J]. Computer Vision and Image Understanding, 2021, 210: 103229. doi: 10.1016/j.cviu.2021.103229.
    [16] BENGIO Y, LOURADOUR J, COLLOBERT R, et al. Curriculum learning[C]. Proceedings of the 26th Annual International Conference on Machine Learning, Montreal, Canada, 2009: 41–48. doi: 10.1145/1553374.1553380.
    [17] ZHAO Yian, LV Wenyu, XU Shangliang, et al. DETRs beat YOLOs on real-time object detection[C]. Proceedings of 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, USA, 2024: 16965–16974. doi: 10.1109/CVPR52733.2024.01605.
    [18] 高鹏, 张若楠, 刘立波. CC-DETR: 基于课程式学习蒸馏的类增量目标检测方法[J]. 物联网学报, 2026. doi: 10.11959/j.issn.2096-3750.WLW25193.

    GAO Peng, ZHANG Ruonan, and LIU Libo. CC-DETR: Class incremental object detection algorithm based on curriculum learning distillation[J]. Chinese Journal on Internet of Things, 2026. doi: 10.11959/j.issn.2096-3750.WLW25193.
    [19] LIU Shilong, ZENG Zhaoyang, REN Tianhe, et al. Grounding DINO: Marrying DINO with grounded pre-training for open-set object detection[C]. Proceedings of the 18th European Conference on Computer Vision, Milan, Italy, 2025. doi: 10.1007/978-3-031-72970-6_3.
    [20] 才华, 冉越, 付强, 等. 多粒度文本感知分层特征交互的视觉定位方法[J]. 电子与信息学报, 2025, 47(11): 4594–4605. doi: 10.11999/JEIT250387.

    CAI Hua, RAN Yue, FU Qiang, et al. Multi-granularity text perception and hierarchical feature interaction method for visual grounding[J]. Journal of Electronics & Information Technology, 2025, 47(11): 4594–4605. doi: 10.11999/JEIT250387.
    [21] PENG Yansong, LI Hebei, WU Peixi, et al. D-FINE: Redefine regression task in DETRs as fine-grained distribution refinement[C]. Proceedings of the International Conference on Learning Representations, Singapore, Singapore, 2025. (查阅网上资料, 无法确认出版地信息是否正确).
    [22] MO Qijie, GAO Yipeng, FU Shenghao, et al. Bridge past and future: Overcoming information asymmetry in incremental object detection[C]. Proceedings of the 18th European Conference on Computer Vision, Milan, Italy, 2025: 463–480. doi: 10.1007/978-3-031-72640-8_26.
    [23] MASANA M, LIU Xialei, TWARDOWSKI B, et al. Class-incremental learning: Survey and performance evaluation on image classification[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(5): 5513–5533. doi: 10.1109/TPAMI.2022.3213473.
    [24] LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft COCO: Common objects in context[C]. Proceedings of the 13th European Conference on Computer Vision, Zurich, Switzerland, 2014: 740–755. doi: 10.1007/978-3-319-10602-1_48.
    [25] ZHANG Lu, WANG Yang, ZHOU Jiaogen, et al. Hierarchical few-shot object detection: Problem, benchmark and method[C]. Proceedings of the 30th ACM International Conference on Multimedia, Lisbon, Portugal, 2022: 2002–2011. doi: 10.1145/3503161.3548412.
    [26] AN Zijia, DIAO Boyu, HUANG Libo, et al. IOR: Inversed objects replay for incremental object detection[C]. Proceedings of 2025 IEEE International Conference on Acoustics, Speech and Signal Processing, Hyderabad, India, 2025: 1–5. doi: 10.1109/ICASSP49660.2025.10888892.
    [27] MA Mingxiao, ZHU Shunyao, and KANG Guoliang. Index-aligned query distillation for transformer-based incremental object detection[EB/OL]. https://arxiv.org/abs/2508.11339, 2025.
    [28] ZHANG Shizhou, LV Xueqiang, XING Yinghui, et al. Revisiting generative replay for class incremental object detection[C]. Proceedings of 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Nashville, USA, 2025: 20340–20349. doi: 10.1109/CVPR52734.2025.01894.
    [29] WANG Jianzhou, WU Yirui, YUAN Lixin, et al. Diffuse&Refine: Intrinsic knowledge generation and aggregation for incremental object detection[C]. Proceedings of the 34th International Joint Conference on Artificial Intelligence, Jeju, South Korea, 2025: 700. doi: 10.24963/ijcai.2025/700.
  • 加载中
图(6) / 表(10)
计量
  • 文章访问数:  17
  • HTML全文浏览量:  3
  • PDF下载量:  2
  • 被引次数: 0
出版历程
  • 修回日期:  2026-11-17
  • 录用日期:  2026-11-17
  • 网络出版日期:  2026-11-21

目录

    /

    返回文章
    返回