Search Results for "localizer"

Detic (“Detecting Twenty-thousand Classes using Image-level Supervision”) is a large-vocabulary object detector that scales beyond fully annotated datasets by leveraging image-level labels. It decouples localization from classification, training a strong box localizer on standard detection data while learning classifiers from weak supervision and large image-tag corpora. A shared region proposal backbone feeds a flexible classification head that can expand to tens of thousands of categories without exhaustive box annotations. The system supports zero- or few-shot extension to novel categories via semantic embeddings and class name supervision, making “open-world” detection practical. ...

MANTI is a one-stop shop N-termini annotation & evaluation solution. MANTI was previously (un)known as muda.pl ahead of v3.7, the project was renamed to MANTI.pl with v3.7 on 2019-06-24. It congregates information from different MaxQuant or DiaNN/MSFragger output files into a master file suitable explicitly for protein neo-termini analyses. The central anchor for the data congregation is the modificationSpecificPeptides.txt or diann-output.pr_matrix.tsv file - additional data is inferred...

-------- ATTENTION START: RENAMING muda.pl was renamed to MANTI.pl with v3.7, project development can be tracked on the MANTI project page on sourceforge.net. Old versions remain here for archival purposes. -------- ATTENTION END muda.pl is an evaluation script (written in Perl) without great dependencies. It congregates information from 4 different MaxQuant output files into a master file suitable explicitly for protein neo-termini analyses. The central anchor for the data...