Jev 驱动的实时 Doom 演示
发布线程展示游戏场景中的实时决策效果。Diogo 说明演示约每秒调用 10 次 Jev,并附有 2 分 43 秒视频。

Jev 中的系统实时解析对话文本,为每句话切换角色面部表情,支持为台词注入情感并处理对方台词的反应,还加入了帧插值和错误修复。
A system in Jev that parses dialogue text in real-time, swapping out character facial expressions for each phrase. It handles both infusing emotion into the lines you speak and reactions to the lines the other character speaks. I've also added frame interpolation, fixed bugs, and