sangminwoo / Explore-And-Match

Official pytorch implementation of "Explore-And-Match: Bridging Proposal-Based and Proposal-Free With Transformer for Sentence Grounding in Videos"

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Regarding Figure 7. on the paper.

vin30731 opened this issue · comments

Hello,
Could you please explain more about fig.7 on the paper?
What does the x-axis and y-axis of the each proposals mean?
Thanks!

Hi, @vin30731!
Thanks for your interest in our work!

A time segment can be represented as <start, end> or <center, width>.
In Figure 7, we showed 10 distributions of time segment predictions, each distribution belongs to a distinct learnable proposal.
Here, x-axis is center and y-axis is width.
If data points are in the top-middle, it means time segments are distributed around the center and their length is long.

I hope this explanation will resolve some ambiguity!

Hi @sangminwoo, thanks for your prompt reply!
So the color of the graph is essentially the same as what y-axis represents?
They both represents the width of the proposal right?
Again, thank you for replying my questions, it is a great work.

You're welcome! Here are the answers for the questions.

So the color of the graph is essentially the same as what y-axis represents? YES!
They both represents the width of the proposal right? YES!