-
In the paper, the equation for NMS Resampling-Linear is as following:
![image](https://user-images.githubusercontent.com/39671748/113708006-4a339780-9713-11eb-9725-b0732eca3024.png)
But, in my opini…
-
Hello author~ I use offline mode to collect the RGB-D information and try to load the depth file from the local file.
I use this code to load the "*.depth.png"
`np.array(Image.open(depth_file).resiz…
-
Hi,
It seems on resize the image quality becomes very poor, this is not a plugin specific issue but more of a canvas issue. Perhaps the following could somehow be included as a custom resampling qua…
-
Simvascular generated centerlines have less resolution (less points) near the inlets and outlets. An option for greater resolution in the centerlines can help in creating post-simulation cross-section…
-
When GeoViews is used with datashader, it can make sense to convert a choropleth into an approximation of the underlying population-based data, as was done for the census dataset here:
![image](htt…
-
https://github.com/dgsmith1988/ECSE-552-Final-Project/blob/9cedc77a11f532e0a7171f5aa4608921b6e9ed81/Code/audio_data_loader.py#L38
It seems as if Librosa already handles things like mono conversion,…
-
Is it possible to resample (to 44.1 or 96 kHz) songs coming from streaming (320Kbps 16bit Spotify) and then process them with DSP for Volumio on Raspberry Pi3 and I2s DAC?
Resampling works but once t…
-
Either use the SDL2 facility, or wrap/corrode `avresample` or `speexdsp`.
-
Upsampling ~3mm or ~4mm BOLD series to 2mm voxels may be excessive. Is it plausible to add a 3mm^3 resampling target?
Related: poldracklab/fmriprep#1360
-
### Describe the issue:
The model I am working with sometimes returns NaN for the logp (it is an ODE model, and for certain nasty parameter combinations the solver gives up). When using SMC as a samp…