LIS1 — Visual Reasoning & Referring Segmentation

Upload an image, load one LIS1 model, and enter up to four object names or natural-language descriptions. Models are downloaded from Hugging Face on demand and loaded in 4-bit quantization.

Model availability: LIS1-9B and LIS1-27B are not available in this Space due to its resource limitations.

LIS1 model size

No model loaded.

0 1

Model forward time: —