Unverified50% confidenceFactExact time
The output of text-to-motion (T2M) is a time series of joint rotations or displacements rather than pixels, so the model's parameter count can be made relatively small.
1
Sources
50%
Confidence
Long-term
Relevance
10/28/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedT2M models typically rely on parametric human body models such as SMPL as the motion representation, and during inference the skeletal motion is retargeted to the target 3D model.68% similarUnverifiedRT-2（2023年）将55B参数的视觉-语言模型PaLM-E微调为可直接输出离散化动作token的策略模型65% similarUnverifiedTIDE does not require visual search or backbone model fine-tuning at inference time; after skill distillation, a single LLM call activates the relevant rules and simultaneously generates positive and negative prompts, then only one T2V generation is performed.63% similarUnverifiedLook-ahead bias is when a model in a quantitative backtest uses information unavailable at time T when making a decision at time T; the solution is bitemporal modeling, allowing only queries for versions whose record time precedes the decision point.62% similarUnverifiedThis experiment directly converted the model's output action probabilities into game operations, rather than generating text.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1109679API
curl https://kongchang.com/api/v1/knowledge/claims/1109679MCP
get_claim(id=1109679)