syamjithnk/ocr-numeral-fidelity
Viewer • Updated • 7k
Syamjith NK
syamjithnk
AI & ML interests
Arabic language technology · speech synthesis evaluation · text rendering and shaping · applied AI in film production
Recent Activity
updated a dataset about 9 hours ago
syamjithnk/ocr-numeral-fidelity published a dataset about 9 hours ago
syamjithnk/ocr-numeral-fidelity updated a dataset about 9 hours ago
syamjithnk/arshapeOrganizations
None yet
updated a
dataset about 9 hours ago
published a
dataset about 9 hours ago
updated 2
datasets about 9 hours ago
Numerals as a stress case for Arabic TTS (open data)
2
#3 opened 2 days ago
by syamjithnk
updated a
collection 2 days ago
updated a
Space 2 days ago
updated a
dataset 2 days ago
updated a
dataset 4 days ago
published a
Space 13 days ago
Labels are stored as presentation forms (shaping applied before saving)
1
#2 opened about 1 month ago
by syamjithnk
published a
dataset about 1 month ago
Arabic-Indic numerals silently break TTS: 7% vs 73% on the same sentences
2
#2 opened about 2 months ago
by syamjithnk
Arabic-Indic numerals collapse to 7% while Western digits hold at 73%
2
#1 opened about 2 months ago
by syamjithnk
ArTST does not speak numerals: 0/30 in digits, and Arabic-Indic digits are absent from the tokenizer vocabulary
#6 opened about 2 months ago
by syamjithnk
Numerals: Arabic-Indic digits are out-of-vocabulary, and Western digits tokenise but are not spoken
#5 opened about 2 months ago
by syamjithnk