Expired50% confidenceFactExact time
One of the prominent pain points of current text-to-video models is the uncontrollability and inconsistency of motion, where the same prompt may generate completely different actions.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/15/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 10/13/2026(expired)
Sources
Related Claims
UnverifiedWhen the semantic direction of text prompts conflicts with the motion characteristics of a reference video, the model must negotiate between two signals, often resulting in unstable motion output.69% similarUnverified坐下、蹲下、弯腰拾物等日常动作容易被误判为跌倒，单帧检测难以区分主动姿态变化与意外跌倒65% similarUnverified研究表明四足运动形态比轮式或双足机器人更容易引发人类情感投射64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/517851API
curl https://kongchang.com/api/v1/knowledge/claims/517851MCP
get_claim(id=517851)