Tagged articles

two-stage training

3 articles · Page 1 of 1

Sep 23, 2026 · Artificial Intelligence

ART: Two-Stage Makeup Transfer Anchors Supervision to Real Images, Beats Pseudo-Target Ceiling

vivo BlueImage Lab introduces ART, a two-stage makeup transfer framework that first learns from synthetic pseudo-targets then refines using real reference images as supervision, achieving state-of-the-art fidelity on complex makeup like glitter and face painting while preserving identity, and releases the first 2K makeup dataset MF2K.

ECCV 2026MF2K datasetcomputer vision

0 likes · 15 min read

ART: Two-Stage Makeup Transfer Anchors Supervision to Real Images, Beats Pseudo-Target Ceiling

Machine Heart

Apr 6, 2026 · Artificial Intelligence

How HumDex Overcomes Humanoid Robot Data Bottlenecks with Low‑Cost Full‑Body Dexterous Control

HumDex combines wireless inertial motion capture, a learning‑based hand‑redirection network, and a two‑stage pre‑training/fine‑tuning pipeline to deliver portable, high‑precision teleoperation for humanoid robots, cutting data‑collection time by 26% and raising remote‑operation success to 91.7% while enabling zero‑shot generalization across unseen objects and scenes.

Humanoid Robotdata collectioninertial motion capture

0 likes · 9 min read

How HumDex Overcomes Humanoid Robot Data Bottlenecks with Low‑Cost Full‑Body Dexterous Control

AI Algorithm Path

Mar 5, 2025 · Artificial Intelligence

Understanding NV-Embed: How NVIDIA’s Decoder‑Only Model Achieves State‑of‑the‑Art Embeddings

This article dissects NVIDIA’s open‑source NV‑Embed model, explaining its decoder‑only architecture, latent attention layer, two‑stage contrastive training, data curation strategies, and experimental results that together push embedding performance to the top of the MTEB benchmark.

EmbeddingMistralNV-Embed

0 likes · 9 min read

Understanding NV-Embed: How NVIDIA’s Decoder‑Only Model Achieves State‑of‑the‑Art Embeddings