Meta announced “Leffa,” an AI image generation technology that can control the virtual fitting and pose of human images. This framework is specialized in human image generation, and its main functions are 1️⃣ [Virtual Fitting] that can virtually change clothes, and 2️⃣ [Pose Control] that can change the pose of the generated person. Leffa provides precise control functions that can finely adjust the appearance and pose of the generated image, and is open source so anyone can use it. --- According to the test results, [Virtual Fitting] generally reflects only the top, and [Pose Control] tends to distort the face. Incomplete.