HomeHomeResearchProjectProjectBlogBlogArticleArticlePaperPaper

Archive

5 entries

Research

Main research in deep learning: representation learning, generative modeling, and neural signal decoding.
Cover
Finished·Sep 2026

Control mapping as a design variable: pricing the posterior-to-command step in motor-imagery BCI

Four ways of turning the same motor-imagery posterior into robot motion, replayed through a simulated planar arm on BCI Competition IV-2a, so the control mapping can be measured against the decoder that feeds it.

EEG decodingMotor imageryDeep learningShared control
Cover
Finished·Jun 2026

Hessian-Guided Feature Injection in U-Net for Nucleus Instance Segmentation

Nucleus instance segmentation in H&E histopathology images using Hessian-derived Morse critical points as structured geometric input features.

Hessian MatricesTopological PriorSaddle PointInstance Segmentation
Cover
Finished·Dec 2025

[ID] Implementasi Arsitektur Mixture of Low-Rank Adapter Experts (X-LoRA) dalam Adaptasi Cross-Lingual Inggris-Indonesia dengan Qwen2.5-0.5B

Membangun Model Bilingual (EN-ID) menggunakan Arsitektur X-LoRA dengan Base Model Qwen2.5-0.5B-Instruct

LLMsPEFTX-LoRACross-Lingual
@syauqi-nabil-tasri000 — Scroll