1. Stylish Woman in Tokyo

A stylish woman walks down a Tokyo street filled with warm glowing neon and animated city signage. She wears a black leather jacket, a long red dress, and black boots, and carries a black purse. She wears sunglasses and red lipstick. She walks confidently and casually. The street is damp and reflective, creating a mirror effect of the colorful lights. Many pedestrians walk about.

2. Thoughtful Man in Paris

An extreme close-up of a gray-haired man with a beard in his 60s, deep in thought pondering the history of the universe as he sits at a cafe in Paris. His eyes focus on people offscreen as they walk, while he sits mostly motionless. He is dressed in a wool coat suit coat with a button-down shirt, wearing a brown beret and glasses, with a very professorial appearance. At the end, he offers a subtle closed-mouth smile as if he found the answer to the mystery of life. The lighting is very cinematic with the golden light and the Parisian streets and city in the background, depth of field, cinematic 35mm film.

3. Chef in the Kitchen

Chef chopping onions in the kitchen for the preparation of the dish.

4. Venetian Canal at Dawn

Through a Venetian canal at dawn, glide as morning light paints ancient facades in pastel hues. Track a solitary gondola as it approaches a weathered stone bridge. Above, wooden shutters open with a creak as an unseen resident scatters crumbs to circling pigeons.

Overview

Goku is a new family of joint image-and-video generation models based on rectified flow Transformers. It is designed to achieve industry-grade performance, integrating advanced techniques for high-quality visual generation, including meticulous data curation, model design, and flow formulation.

Key Contributions

• 📊 High-quality fine-grained image and video data curation.
• 🔄 The pioneering use of rectified flow for enhanced interaction among video and image tokens.
• 🌟 Superior qualitative and quantitative performance in both image and video generation tasks.

Supported Generation Tasks

• 🎬 Text-to-Video Generation
• 🖼️ Image-to-Video Generation
• 🎨 Text-to-Image Generation

Performance Benchmarks

Goku achieves top scores on major benchmarks:

• 0.76 on GenEval (text-to-image generation)
• 83.65 on DPG-Bench (text-to-image generation)
• 84.85 on VBench (text-to-video generation)

VBench Performance

Goku-T2V achieves an impressive score of 84.85 in VBench, securing the No.2 position as of 2024-10-07, surpassing several leading commercial text-to-video models.