-
hi
I am very impressive on the speed of Clickhouse, greet job
Now, I want to add the feature search to clickhouse, search the nearest feature by euclidean distance. Oue feature is 512d float32, …
-
Print the nearest neighbors for these words and send by email (sort by confidence): apple, mango, BMW, USA, Berlin, IBM, jaguar. The candidates are ALL words in the vocabulary of the pre-trained hypen…
-
Use case: I'm trying to set my background to a 64x36 pixel art image, but the image is heavily blurred. I've since set it to an upscaled version but this seems inefficient.
Feature being requested:…
-
The minimum value for n_neighbors param in umap.UMAP() is incorrect based on the value error in the code below. It returns `ValueError: n_neighbors must be greater than 2` if entering 1, but 2 is a va…
-
In get batch, the images seem to be re-sized using a nearest-neighbor interpolation,
look at line 77:
ims(oky,okx,:,si) = bsxfun(@minus, rgb(sy(oky),sx(okx),:), opts.rgbMean) ;
Is this on purp…
-
not really an issue, but a question/comment as id love to contribute to phistoffury... have you looked into piggybacking off spotify's annoy framework in lieu of kd trees?
https://github.com/spotif…
-
We should remove the pages copied from the earlier version of the Berkeley textbook - to simplify copyright.
Plan:
* Remove Literary Characters section
* Remove Classifiers section
* Search fo…
-
I understand the ground truth frame is generated by taking the two nearest neighbors of each atom. How is the predicted frame generated? Do you perform two nearest neighbors on the predicted coords?
-
It looks like behavior might have changed from stable to the alpha regarding large images being downsampled. The case below is a bit of an extreme one, but a 3200x3200 image resized to 200x200 previou…
mnutt updated
3 years ago
-
http://stackoverflow.com/questions/5696859/how-do-i-use-kd-trees-for-determining-string-similarity