Md. Asaduzzaman Shuvo

Datasets

Public datasets and benchmarks released with my papers.

  1. [D1]WiNLP@EMNLP 2026
    BanglaCEH:- When a Name Is Not a Name: A Benchmark Dataset and Distilled Reasoning for Culturally Entangled Bangla Homographs in Low-Resource LLMs.

    Hugging Face · NLP · Released with EMNLP 2026

    • 1.52k samples
    • CC BY 4.0
    DatasetCodePaper
  2. [D2]GroundLM@EMNLP 2026
    BLADE:- Polite on the Surface, Broken in Practice: A Curated Dataset for Fixing Generation and Register Failures in Low-Resource Bangla Text Generation

    Hugging Face · NLP · Released with EMNLP 2026

    • 4.2K samples
    • CC BY 4.0
    DatasetCodePaper
  3. [D3]AI4RWC@CVPR 2026
    VisText-Mosquito: A Multimodal Dataset for Mosquito Breeding Site Detection, Surface Segmentation, and Reasoning

    Mendeley Data · Computer Vision, Object Detection · Released with CVPR 2026

    • 1.85k images
    • CC BY 4.0
    DatasetCodePaper
  4. [D4]arxiv
    WADE: A Waste Annotation and Reasoning Chain Dataset for Waterbody Monitoring in Bangladesh

    Mendeley Data · Computer Vision, Object Detection, Reasoning · Released with arxiv

    • 2.20k images
    • CC BY 4.0
    DatasetPaper