5 papers
M3DDM+: An improved video outpainting by a modified masking strategy
Takuya Murakawa, Takumi Fukuzawa, Ning Ding +1
M3DDM provides a computationally efficient framework for video outpainting via latent diffusion modeling. However, it exhibits significant quality degradation -- manifested as spat…
Shot2Tactic-Caption: Multi-Scale Captioning of Badminton Videos for Tactical Understanding
Ning Ding, Keisuke Fujii, Toru Tamaki
Tactical understanding in badminton involves interpreting not only individual actions but also how tactics are dynamically executed over time. In this paper, we propose \textbf{Sho…
Disentangling Static and Dynamic Information for Reducing Static Bias in Action Recognition
Masato Kobayashi, Ning Ding, Toru Tamaki
Action recognition models rely excessively on static cues rather than dynamic human motion, which is known as static bias. This bias leads to poor performance in real-world applica…
MoExDA: Domain Adaptation for Edge-based Action Recognition
Takuya Sugimoto, Ning Ding, Toru Tamaki
Modern action recognition models suffer from static bias, leading to reduced generalization performance. In this paper, we propose MoExDA, a lightweight domain adaptation between R…
Separating Shared and Domain-Specific LoRAs for Multi-Domain Learning
Yusaku Takama, Ning Ding, Tatsuya Yokota +1
Existing architectures of multi-domain learning have two types of adapters: shared LoRA for all domains and domain-specific LoRA for each particular domain. However, it remains unc…