Method Drift›Parameter-efficient fine-tuning (LoRA family)
Superseded baseline#1012 of 1,113 most-superseded
SVFormer (Rizvi et al. 2024)
Parameter-efficient fine-tuning (LoRA family)
superseded — cited as a baseline and beaten by newer methods
1 papers critique it · 0 beat it on benchmarks
What papers say
Verbatim critique sentences, each from a paper that cites SVFormer (Rizvi et al. 2024) as a baseline.
“Both of these works are promising, but they also show how challenging such fine-gran activity recognition is, and that there is improvement potential for this application.”
— Can Local Vision-Language Models improve Activity Recognition over Vision Transformers? -- Case Study on Newborn Resuscitation
Newer alternatives
Recent methods in the same sub-problem, not yet superseded in the knowledge base.