PulseAugur
EN
LIVE 08:50:42

Compact AI model achieves benchmark saturation for Devanagari recognition

By PulseAugur Editorial · [3 sources] ·

Researchers have developed a highly efficient convolutional neural network, named Barnamala, for recognizing handwritten Devanagari script. This compact model, with only 1.11 million parameters, achieves a benchmark saturation accuracy of 99.73% on the DHCD dataset, significantly outperforming larger models in terms of size and efficiency. Barnamala also demonstrates strong zero-shot performance on the CMATERdb dataset and superior robustness against data corruption compared to its larger counterparts. AI

IMPACT Demonstrates significant efficiency gains in AI model design for character recognition tasks.

RANK_REASON The cluster contains an academic paper detailing a new model and its performance on specific benchmarks.

Read on arXiv cs.AI →

AI-generated summary · Google Gemini · from 3 sources. How we write summaries →

Compact AI model achieves benchmark saturation for Devanagari recognition

COVERAGE [3]

  1. arXiv cs.AI TIER_1 English(EN) · Ashish Thapa, Samrat Karki ·

    Barnamala: Parameter-Efficient Handwritten Devanagari Recognition at Benchmark Saturation

    arXiv:2607.13689v1 Announce Type: cross Abstract: We built a compact convolutional network (1.11 M parameters) for 46-class DHCD Devanagari recognition and reached 99.73%, the highest reported at 15.6x smaller than prior state-of-the-art. We have effectively reached the saturatio…

  2. arXiv cs.AI TIER_1 English(EN) · Samrat Karki ·

    Barnamala: Parameter-Efficient Handwritten Devanagari Recognition at Benchmark Saturation

    We built a compact convolutional network (1.11 M parameters) for 46-class DHCD Devanagari recognition and reached 99.73%, the highest reported at 15.6x smaller than prior state-of-the-art. We have effectively reached the saturation point: every model tested, large teacher ensembl…

  3. Hugging Face Daily Papers TIER_1 English(EN) ·

    Barnamala: Parameter-Efficient Handwritten Devanagari Recognition at Benchmark Saturation

    We built a compact convolutional network (1.11 M parameters) for 46-class DHCD Devanagari recognition and reached 99.73%, the highest reported at 15.6x smaller than prior state-of-the-art. We have effectively reached the saturation point: every model tested, large teacher ensembl…

RELATED ENTITIES

RELATED TOPICS