套磁6封，一封都没回，求分析

lz是zju的本科，sjtu的小硕，专业自动化模式识别，准备申请computer vision 方向，10月中旬才开始套，套了几个前50的学校老师，没人回复%>_<%
发一封套磁的信，求大家分析一下：

Dear Professor ***

I am currently a Master student at Shanghai Jiao Tong University,
China. I am considering applying to **’s PHD program and would be
very interested joining your research group.

I have read your recent papers “Understanding scenes on Many Levels”
and your paper on “Spatial Pyramid Matching”. Since I’m also working on
visual recognition and related fields, I find your research very interesting
and want to briefly talk about some of my ideas on it:

Firstly, about the differences between Spatial Pyramid Matching and Pyramid
Matching Kernels. Although Spatial Pyramid Matching is based on the idea
of Pyramid Matching Kernels, their working schemes are totally different:
PMK tries to capture the similarity of variable-sized feature sets in
every level, while SPM tends to work in a statistical fashion to extract the
spatial information of feature.

On the paper “Understanding scenes on Many Levels” , the key idea
is to introduce multiple types of labels instead of a single label, and
to leverage the consistency between them. I think that maybe we
could do the parsing and categorization work together, or use one to
help the other. For example, if we could find out that one region is an
animal, then the image would be more likely to categorized into an animal
class. It seems that it will help but I didn’t do the experiments.

For my Ph.D application, I am wondering whether you have open positions for
PhD students next year. My CV is attached for your review. I would
greatly appreciate you if you could give me a few words at your
convenience. Thanks.

Best Regards,
**