Google's Gemini Omni 1.2 Flash extends video generation to 15 seconds with iterative editing, enabling workflows like selfie street vlogs, anime cooking clips, 2D-to-3D character integration, motion transfer, and multi-segment educational videos, while pricing drops to $0.07/second for 720p API access.
PhiZero introduces a compact “physical language” representation that lets world models generate realistic physics‑driven videos, transfer motions across bodies, and predict interactive scenes, demonstrating superior performance on generation and understanding benchmarks compared to pixel‑based approaches.
Pixelle-Video lets users input a single topic and, within minutes, automatically generates a short video with script, AI‑generated visuals, synthesized voice‑over, background music, and final rendering, eliminating the need for manual editing or specialized software.
The Deformable Anchor Model (DAM) and its hierarchical extension (HDAM) enhance image‑driven motion transfer by introducing a learned root (and intermediate) node as a structural prior for keypoint detection, yielding clearer motions, lower reconstruction errors across multiple benchmarks, while preserving the speed of the original First‑Order Motion Model.
