MovingDroneCrowd VIC & Tracking
Count & track every person in moving-drone crowd video
None defined yet.
Count & track every person in moving-drone crowd video
Zero-shot voice cloning with the UniVoice ASR+TTS LLM
190K-param sentiment classifier (Neg/Neu/Pos)
Late-interaction visual document retrieval with NeoMME
Passage QA with a 7.5M-param CPU-trained LM
Text to knowledge graph in one non-autoregressive pass
Multi-subject reference images to video with LTX-2.5
190K-param sentiment classifier (Neg/Neu/Pos)
Structured MiniMax-H3 audio-video prompts from short ideas
Persian text-to-speech with voice cloning, on CPU
Localize hallucinated spans against evidence
Large-baseline novel view synthesis from one image
Drive a camera, press F, and the character greets you
Detect hard cuts and gradual shot transitions in video
Anime-style Japanese TTS with voice cloning
Surgical concept erasure in CogVideoX-5b video diffusion
Adaptive-compute ViT that classifies in progressive rounds
Feed-forward 3D Gaussian Splatting from unposed video
Accelerated masked image generation with Lumina-DiMOO
Generate instruction-aligned background music for any video
Sequence-only protein-ligand binding scorer (LULA-1.1)
Forecast a future brain MRI from a baseline scan
Segment any category you can name, in a single pass
Segment 11 brain structures in ultra-low-field MRI