The second evaluation campaign, organized by the MIPRCV consortium, comprises a set of benchmarks where the tasks have been defined with the goal of evaluating and comparing new techniques closely related to the main interests of the project. We invite the members of the consortium to participate in this second campaign. There will be six main tasks, some of them as a continuation of the first campaign and new ones.

• MNIST database of handwritten digits

Second edition of this benchmark of handwritten digits. In this second campaign the goal is to apply active and online learning methods. Organizer: Ernest Valveny (PR-CVC group)

• Image Relevance Feedback benchmark

Second edition of this benchmark where the goal is, as in the first campaign, to compare different strategies of user interaction in the image relevance feedback context. Organizer: Roberto Paredes (PRHLT-ITI group)

• BiModal HTR

Second edition of this benchmark of on-line and off-line handwritten text. The challenge is to use multimodal fusion techniques to classify each pair of on/off-line test samples with the highest accuracy. Organizers: Moisés Pastor and Enrique Vidal  (PRHLT-ITI group)

• Multidimensional Face benchmark

Second edition of this benchmark of faces. The goal is the classification of human faces involving simultaneously various class labels: gender, race, age and facial expression. In this edition, new auxiliar class labels with dependency have been introduced to improve the classification. Organizer: Luis Baumela (CVL-UGR group)


• The Karyo benchmark

This is a new benchmark to test and develop interactive protocols. It contains karyotypes, where each one is composed of 22 chromosome images. The goal is to associate each chromosome image with a label from a set of 22. Organizers: Veronica Romero and Vicente Alabau (PRHLT-ITI/UPV group)

• Video retrieval benchmark
  

New bechmark extracted from a set of videos of the TRECVID 2007 video retrieval set. These videos were divided into shots, and a keyframe was selected for each shot. This benchmark will be used for information retrieval, as the image relevance feedback benchmark. Organizer: Ángeles López (CVDSP-UJI group)