Hello, I would like to inquire about the synthetic datasets. about ps-fcn HOT 5 CLOSED

Sorry, in addition, the merl dataset at the website https://www.merl.com/brdf/ is no longer available for download. Can you provide merl data set download link? Thank you very much for your reply!

from ps-fcn.

Hi,

Although I set the random seed to 0 during rendering, it might not be possible to render exactly the same dataset might not be possible.

Please see https://drive.google.com/file/d/19fUBvRHWklg_vHLT8wuhL36ZmtYc7jOx/view?usp=sharing for a copy of MERL dataset.

A sample .xml file can be found in the following
`

<integrator type="multichannel">
    <integrator type="field">
        <string name ="field" value="shNormal" />
        <spectrum name="undefined" value="0" />
        <!--Occlusion-->
    </integrator>
</integrator>

<shape type="sphere" id="obj1">
    <point name="center" x="$tx" y="$ty" z="$tz" />
    <float name="radius" value="$scale" />
</shape>

<sensor type="orthographic">
    <float name="farClip" value="55.864"/>
	<float name="nearClip" value="0.10864"/>
	<transform name="toWorld">
        <lookat target="0, 0, -14" origin="0, 0, -15" up="0, 1, 0"/>
	</transform>

	<sampler type="ldsampler">
		<integer name="sampleCount" value="64"/>
	</sampler>

	<film type="hdrfilm">
		<integer name="height" value="$h"/>
		<integer name="width" value="$w"/>

        <string  name="pixelFormat"  value="rgb"/>
        <string  name="channelNames" value="normal"/>
		<rfilter type="gaussian"/>
        <boolean name="banner" value="false"/>
	</film>
</sensor>

from ps-fcn.

Thank you very much for your reply. your guidance is very helpful to me.

So I combine this XML sample file with blobby dataset, merl dataset, sculpture dataset and MITSUBA renderer. Can I render the Synthetic datasets described in this article?

Because I noticed that the amount of data is very large, I downloaded it and unzipped it and it took me half a day, hhh.it may be troublesome to render manually.

So I'm sorry to ask if there is a faster way to get this synthetic data set? The data set does not need to be as large as described in the paper, it is just a relatively small data set.

Thank you very much again for your reply!

from ps-fcn.

Oh, you might be using a remote internet disk. In my local hard disk, it takes around one hour to unzip the data.
I think you can just use a subset of the dataset. It might not be easy to render the dataset by yourself within a short time .

from ps-fcn.

Thank you for your reply, your suggestion is very helpful to me!

from ps-fcn.