We present a further analysis of visual modality incompleteness, benchmarking latest MMEA models on our proposed dataset MMEA-UMVM.
5 PAPERS • 7 BENCHMARKS