Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jan 17, 2026

Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody
09:09

Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody

Published on: September 27, 2024

814

Semi-Supervised Text-Based Person Search.

Daming Gao, Yang Bai, Min Cao

    IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
    |September 16, 2025
    PubMed
    Summary
    This summary is machine-generated.

    Related Concept Videos

    Sign Test for Matched Pairs01:17

    Sign Test for Matched Pairs

    383
    The sign test for matched pairs offers a robust method for comparing two paired samples, often for the effects of an intervention in one of them. This method is very useful in situations where the underlying distribution of the data is unknown. The test compares two related samples—often pre- and post-treatment measurements on the same subjects—to determine if there are significant differences in their median values.
    To conduct the sign test, we first calculate the differences in...
    383

    You might also read

    Related Articles

    Articles linked to this work by shared authors, journal, and citation graph.

    Sort by
    Same author

    LoRASculpt: Harmonious Low-Rank Adaptation for Multimodal Large Language Models.

    IEEE transactions on pattern analysis and machine intelligence·2026
    Same author

    Identification of <i>LiMYC</i> and <i>LiTPS</i> Gene Families Involved in MeJA-Induced Terpene Accumulation in <i>Lagerstroemia indica</i> 'Whit III'.

    Plants (Basel, Switzerland)·2026
    Same author

    HiSymGeo: Hierarchical Context Symbiosis for Cross-View Object-Level Image Geo-Localization.

    IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
    Same author

    A high-resolution dataset on costs and greenhouse gas emissions of battery recycling in China.

    Scientific data·2026
    Same author

    Molecular glue degraders of HuR suppress BRAF-mutant colorectal cancer.

    Nature·2026
    Same author

    Hippo Signaling Suppresses Cell Ploidy and Tumorigenesis through Skp2.

    Cancer cell·2026
    Same journal

    Mask-guided Asymmetric Contrastive and Semantic Alignment for Unsupervised Person Re-Identification.

    IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
    Same journal

    Hyperbolic Cycle Alignment for Infrared-Visible Image Fusion.

    IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
    Same journal

    Learning Gaze Synthesizer via 3D-eye Controlled Diffusion and Cross-domain Feature Alignment.

    IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
    Same journal

    Underlying Semantic Diffusion for Effective and Efficient In-Context Learning.

    IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
    Same journal

    DiffRES: Unleashing Text-to-Image Diffusion Models for Generative Referring Expression Segmentation without Information Leakage.

    IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
    Same journal

    Location Matters: Frequency-Spatial Dual Space Adaptation for Cross-Domain Few-Shot Segmentation.

    IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
    See all related articles

    This study introduces a novel semi-supervised approach for text-based person search (TBPS), overcoming data annotation challenges. The proposed framework effectively handles noisy generated data, improving retrieval accuracy with limited annotations.

    Area of Science:

    • Computer Science
    • Artificial Intelligence
    • Machine Learning

    Background:

    • Text-based person search (TBPS) typically requires extensive annotated image-text data for fully-supervised learning.
    • Acquiring detailed textual descriptions for large datasets of person images is practically challenging and resource-intensive.

    Purpose of the Study:

    • To explore and develop effective methods for TBPS in a semi-supervised setting with limited annotated data.
    • To address the performance limitations of existing TBPS methods due to scarce annotated image-text pairs.

    Main Methods:

    • A two-stage generation-then-retrieval framework is proposed, starting with generating pseudo-texts for unannotated images using image captioning.
    • A noise-robust retrieval framework is introduced, incorporating Hybrid Patch-Channel Masking (PC-Mask) and Noise-Guided Progressive Training (NP-Train).

    Related Experiment Videos

    Last Updated: Jan 17, 2026

    Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody
    09:09

    Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody

    Published on: September 27, 2024

    814
  • PC-Mask refines model architecture by masking input data at patch and channel levels to mitigate overfitting on noisy pseudo-texts.
  • NP-Train enhances training by progressively scheduling based on pseudo-text noise levels for robust learning.

    • Main Results:

    • The proposed semi-supervised TBPS framework demonstrates promising performance across multiple benchmarks.
    • The noise-robust strategies (PC-Mask and NP-Train) significantly improve the model's ability to handle noisy pseudo-text data.
    • The generation-then-retrieval approach effectively leverages limited annotated data and generated pseudo-texts for improved retrieval.

    Conclusions:

    • Semi-supervised learning is a viable and effective approach for TBPS, significantly reducing the reliance on fully annotated datasets.
    • The developed noise-robust retrieval framework offers a practical solution for improving TBPS performance in low-annotation scenarios.
    • This research paves the way for more scalable and efficient TBPS systems by addressing data scarcity challenges.