--- title: Organization Card sdk: static colorFrom: indigo colorTo: gray --- # Articulated Research Institute for Scientific Excellence **A.R.i.S.E.** is an applied machine learning research lab based in Bangladesh. We work wherever rigorous ML methodology meets real-world complexity โ€” from financial markets and satellite imagery to regional languages, medical imaging, explainability, and beyond. Our research spans the full arc from early-stage ideas to published results. Some projects are foundational explorations; others carry peer-reviewed contributions and public datasets. What ties them together is a commitment to methodologically sound, reproducible science โ€” with a focus on problems that matter in South Asian and low-resource contexts. We are affiliated with the **Department of Computer Science and Engineering, Southeast University, Bangladesh**. ๐ŸŒ [ariserl.org](https://ariserl.org)  ยท  ๐Ÿ’ผ [LinkedIn](https://bd.linkedin.com/company/ariserl24)  ยท  ๐Ÿ“˜ [Facebook](https://www.facebook.com/ariserl24) --- ## What We Work On Our portfolio is intentionally broad. Past and ongoing work has touched: - **Financial ML** โ€” structural break detection, coverage bias, regime analysis, and time series forecasting in emerging markets - **Remote Sensing** โ€” satellite and aerial image analysis - **Regional Language NLP** โ€” dialect-aware datasets and models for Bengali and related varieties - **Medical & Biomedical Imaging** โ€” retinal disease detection, ocular health, and physics-informed vision - **Explainable AI** โ€” gradient-based and attention-based interpretability on image models - **Multimodal & Fusion Architectures** โ€” combining heterogeneous inputs for robust prediction This list grows. If something is in our pipeline, it will eventually surface here. --- ## What You'll Find Here - **Datasets** โ€” annotated and benchmark datasets released alongside our publications - **Model Checkpoints** โ€” fine-tuned and adapter-based models from our research pipelines - **Reproducibility Artifacts** โ€” configs, notebooks, and scripts to reproduce key results --- ## Get in Touch We welcome collaboration with researchers working on applied ML problems, particularly in low-resource, domain-specific, or underrepresented settings. Reach us at [ariserl.org](https://ariserl.org) or connect on [LinkedIn](https://bd.linkedin.com/company/ariserl24).