LifeCLEF Plant Identification Task 2014

2025-09-28

AI Summary

This study addresses fine-grained identification of 500 tree and herbaceous plant species under real-world conditions. To this end, we introduce the first large-scale, multi-view plant image benchmark dataset derived from a citizen science initiative, comprising six uncontrolled, in-the-wild image modalities: leaf scans, flowers, fruits, bark, twigs, and whole-plant photographs. We propose a unified recognition framework integrating multi-source visual features and cross-modal retrieval, supporting fine-grained classification, multi-view feature alignment, and large-scale image retrieval. Twenty-seven methods from ten teams across six countries participated in the benchmark evaluation, validating the dataset’s ecological validity, taxonomic coverage, and imaging diversity. Our work significantly enhances robustness and generalization of plant identification in complex natural settings, establishing a high-value data foundation and methodological paradigm for biodiversity monitoring and interpretable plant AI.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Security and Privacy: Data transparency and provenance
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

The LifeCLEFs plant identification task provides a testbed for a system-oriented evaluation of plant identification about 500 species trees and herbaceous plants. Seven types of image content are considered: scan and scan-like pictures of leaf, and 6 kinds of detailed views with un- constrained conditions, directly photographed on the plant: flower, fruit, stem & bark, branch, leaf and entire view. The main originality of this data is that it was specifically built through a citizen sciences initiative conducted by Tela Botanica, a French social network of amateur and expert botanists. This makes the task closer to the conditions of a real- world application. This overview presents more precisely the resources and assessments of task, summarizes the retrieval approaches employed by the participating groups, and provides an analysis of the main eval- uation results. With a total of ten groups from six countries and with a total of twenty seven submitted runs, involving distinct and original methods, this fourth year task confirms Image & Multimedia Retrieval community interest for biodiversity and botany, and highlights further challenging studies in plant identification.

Problem

• Identify 500 plant species using diverse images
• Evaluate plant recognition under real-world conditions
• Analyze citizen-science collected botanical image data

Innovation

• Combined citizen science data collection
• Used seven plant image types
• Evaluated twenty seven retrieval methods