Peng Liu

The paper introduces GLOBE, a brain-wide single-cell recording technology, and demonstrates its use in capturing Fos transcriptional activity across a mouse brain over time.

This paper introduces a graph-based framework for correcting nucleus segmentation errors in 3D reconstructions, achieving high accuracy on brain-scale benchmarks.

The paper develops NucleoNet and DropNet, deep learning models for segmenting nuclei and lipid droplets in EM images, and applies them to compare 2D/3D cancer models with in vivo tumors.

MitoEM 2.0 provides a benchmark for 3D mitochondria segmentation in EM images with diverse datasets and standardized labels for model evaluation and comparison.

WormID-Bench evaluates methods for neuron detection, identification, and tracking in C. elegans using a multi-laboratory dataset, highlighting the need for improved computational approaches.