-
Hi,
In the demos it shows, that you can also extract the type of the segment (e.g a door, a human etc.) can it be extracted from the model or do we need an additional library for that?
Cheers
-
hi, for the txt image genreateion, why not try using existed LLMs, plus the decoder in tokenizer, and training the whole model with lora?
just like SEEd does.
So that it at least won't harm the …
-
Does every masks have a semantic classes group index or name?
-
Thanks for leasing this wonderful work!
I saw the demo shows examples of using point, box as input prompt. Does the demo support text as prompt?
-
Hello,
I am testing the performance on some sample images from ADE20k (e.g. ADE_test_00000002.jpg, ADE_test_00000005.jpg). I am using the vocabulary from "cls_ade20k.txt" for query.
I first che…
-
Utterances can point to tiers in a meronymic relationship, but the meaning is "U has subpart of type tier T".
Words can point to tiers in a meronymic relationship, but the meaning is "W is itself pa…
-
[The format of the issue]
Paper name/title:
Paper link:
Code link:
amusi updated
2 months ago
-
Hi Mathieu,
Hope you are doing well !
I created another tool (httpRequest) for find object based on Tufao web server (https://github.com/vinipsmaker/tufao/tree/0.x, as Find-Object is using QT 4.8) …
ghost updated
8 years ago
-
Hi,
I have implemented text prompt-controlled segmentation using selective search and CLIP. Can you suggest any additional techniques I can include? I am considering trying CLIP-GradCAM #4
h…
-
Email: Sofia Holke
Projektledare • It-utveckling • Riksdagsförvaltningen
Telefon: 08-786 43 83 • Mobil: 076-721 65 54 • E-post: [sofia.holke@riksdagen.se](mailto:sofia.holke@riksdagen.se)
[Linked …