A. Retrieval EvaluationFirst, the l

A. Retrieval Evaluation
First, the low dimensional subspaces of the cartoon
database are learned by the subspace learning algorithms,
including semi-MSL, MSL, UMSL, UBDML, CGS,
FCSL, DMSL, and SSL, and second, based on each lowdimensional
subspace, a standard retrieval procedure is
conducted for all characters in the database. In detail, for
each category, one cartoon character is selected as a
query, and thus, 100 cartoon characters are used as
queries, respectively. Then, all the other characters in the
database (including other categories) are ranked according
to the Euclidean distance to the query computed in the
low-dimensional subspace. For each query cartoon
character, the top N characters are obtained, among which
the number of characters from the same class as the query
shape is recorded. The precision rate, shown in table 1 is
the ratio of the recorded number to N.

ผลลัพธ์ (อังกฤษ) 2:

A. Retrieval Evaluation
First, the low dimensional subspaces of the Cartoon
Database are learned by the Learning subspace algorithms,
including semi-MSL, MSL, UMSL, UBDML, CGS,
FCSL, DMSL, and SSL, and second, based on each Lowdimensional
subspace. , a standard Retrieval procedure is
conducted for all characters in the Database. In detail, for
each Category, Cartoon Character one is selected as a
query, and thus, 100 Cartoon characters are used as
queries, respectively. Then, all the characters in the Other
Database (including Other categories) are ranked according
to the Euclidean Distance to the query computed in the
low-dimensional subspace. For each query Cartoon
Character, the top N characters are obtained, among which
the Number of characters from the Same Class as the query
Shape is Recorded. The Precision rate, shown in Table 1 is
the ratio of the Recorded Number to N.