Mitigating Noisy Correspondence in Video–Text Retrieval via Noise-Mined Adaptive Self-Labeling
Noise-mined Adaptive Self-Labeling (NASL), which effectively manages noisy data during training and improves the separation between clean and noisy data while effectively mitigating noise, leading to significant performance improvements on the MSR-VTT, DiDeMo, MSVD, LSMDC, and ActivityNet datasets.