Welcometo thePartial Matching Trackon Watertight Modelsof SHape REtrieval Contest 2007 |
SHREC '07 is a competition whose objective is to
evaluate the effectiveness of 3D-shape retrieval algorithms. This year the contest is divided into different
tracks. This page refers to the Partial Matching track.
Two collections of 3D models will be provided to the participants of the track: the data-set and the query-set:
Watertight Models are object models represented by seamless surfaces, i. e. without defective holes or gaps. Each query model shares common subparts with (possibly) more than one model belonging to the data-set. Given a query model, the goal of this track is the retrieval of the largest number of models belonging to the data-set, such that they share similar subparts with the given query.
A set of relevant, marginally relevant and not relevant models (belonging to the data-set) is associated to each query model.
This set (the ground-truth) is defined a priori by the organizers of the track and will be used to evaluate the performance
of the retrieval algorithms. The performance will be evaluated by using the measures and tools developed in SHREC '06.
Results are now ready. Here you can find all the most significant graphs we extracted.
Given a database of 400 watertight models, and a set of 30 hybrid query models, we defined for each query the set of
highly and marginally relevant items, that is, respectively, the classes which share a similar subpart with the query,
and those ones which are reasonably similar to it.
Dataset | Queryset |
Query Number | Relevant Classes | Marginally Relevant Classes |
1 | cup, teddy | vase, four legs |
2 | human, table | armadillo, chair |
3 | buste, mechanic |   |
4 | plier, spring | airplane, bird |
5 | ant, glasses | octopus |
6 | four legs, airplane | bird, plier, teddy |
7 | armadillo, vase, bearing | human, cup |
8 | fish, bird, mechanic | airplane, plier |
9 | chairs, bearings | tables |
10 | human, table | armadillo, chair |
11 | fisher, hand |   |
12 | human, octopus | armadillo, ant |
13 | hand, spring |   |
14 | human, fish | armadillo |
15 | four legs, vase | cup, teddy |
16 | bird, buste | airplane, plier |
17 | chair, plier | airplane, bird, table |
18 | ant, octopus |   |
19 | airplane, armadillo | human, bird, plier |
20 | teddy, spectacle | four leg |
21 | cup, springs | vase |
22 | four legs, cup | vase, teddy |
23 | armadillo, bearing, bird | human, airplane, plier |
24 | airplane, bird | plier |
25 | head, vase | cup |
26 | chair, table |   |
27 | teddy, hand | four legs |
28 | octopus, plier | bird, airplane, ant |
29 | airplane, mechanical | bird, plier |
30 | four legs, human | armadillo, teddy |
Here you can find all sort of graphs we computed, and a "readme" file which explains what is
in each folder.
Last update: 08/06/2007