The imageCLEF plant identification task 2013

Abstract

Abstract. The ImageCLEF's plant identification task provides a testbed for the system-oriented evaluation of plant identification, more precisely on the 126 tree species identification based on leaf images. Three types of image content are considered: Scan, Scan-like (leaf photographs with a white uniform background), and Photograph (unconstrained leaf with natural background). 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 evaluation results. With a total of eleven groups from eight countries and with a total of 30 runs submitted, involving distinct and original methods, this second year pilot task confirms Image Retrieval community interest for biodiversity and botany, and highlights further challenging studies in plant identification.

Plant identification: man vs. machine

“…A total of 10 participating groups submitted 27 runs that are summarised in the overview working note of the task [19] and further developed in the individual working notes of the participants who submitted one (BME TMIT [22], FINKI [28], I3S [31], IBM AU [26], IV-Processing [30], MIRACL [23], PlantNet [24], QUT [25], Sabanki-Okan [27], SZTE [29]). Official results of the evaluation (based on the score S detailed above) are synthesized in the overview paper of LifeCLEF lab [32] and further developed in the working note of the plant task [19].…”

“…Official results of the evaluation (based on the score S detailed above) are synthesized in the overview paper of LifeCLEF lab [32] and further developed in the working note of the plant task [19]. For the purpose of this paper, we refined the analysis of the collected runs in order to assess the impact of the quality of the test pictures on the identification performances of the evaluated systems.…”

“…It is noticeable that most image-based identification methods and evaluation data proposed in the past were so far based on leaf images (e.g. in [13,20,21] or in the more recent methods evaluated in [17]). Only few of them were focused on images of flower as in [6] or [33].…”

“…A related initiative is the plant identification evaluation task organized since 2011 in the context of the international evaluation forum CLEF 2 ; the task is based on the data collected within Pl@ntNet. In 2011, 2012 and 2013 respectively 8, 11 and 12 international research groups participated in this large collaborative evaluation by benchmarking their image-based plant identification systems (see [16,17,18] for more details). The data used during these 3 first years can be accessed online 3 .…”

“…74 research groups worldwide registered and 10 of them crossed the finish line by submitting runs (from 1 to 4 depending on the teams). Details of the participants and the methods used in the runs are synthesised in the overview working note of the task [19]. The main contribution of this new paper is to complete the system-oriented evaluation of LifeCLEF by a user trial aimed at comparing the performances of state-of-the-art automatic recognition systems with the human expertise (see section 3).…”