SCFAE
No mentions found
This entity hasn't been tracked yet, or Iris is still building its knowledge base.
Related Articles from SNS
Multimodal Fusion via Self-Consistent Task-Gradient Fields
arXiv:2410.15475v2 Announce Type: replace Abstract: Multimodal learning aims to preserve as much task-related information as possible from different inputs. However, current fusion designs often distort the feedback loop to feature extractors. Aggressively merging modalities entangles their representations, making the feature extractors fragile to incomplete inputs.