Video Summarization on TVSum (Correlation Metrics)
0.217Kendall's TauTripleSumm (MoSu)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| TripleSumm (MoSu)training=pre-trained on MoSu and fine-tuned on target dataset2026.03 | 0.217 | 0.282 | |
| TripleSumm (MoSu)Training=Pre-trained on MoSu and fine-tuned on target dataset2026.03 | 0.217 | 0.282 | |
| LLMVSModality=Visual + Text2025.04 | 0.211 | 0.275 | |
| TripleSumm (Full)training=trained on target dataset2026.03 | 0.211 | 0.275 | |
| TripleSumm (Full)Training=Trained on target dataset2026.03 | 0.211 | 0.275 | |
| PGL-SUMModality=Visual2025.04 | 0.206 | 0.157 | |
| DMASumModality=Visual2025.04 | 0.203 | 0.267 | |
| TripleSumm (MoSu)training=pre-trained on MoSu and fine-tuned on target dataset2026.03 | 0.2 | 0.262 | |
| TripleSumm (MoSu)Training=Pre-trained on MoSu and fine-tuned on target dataset2026.03 | 0.2 | 0.262 | |
| TripleSumm (Full)training=trained on target dataset2026.03 | 0.198 | 0.259 | |
| TripleSumm (Full)Training=Trained on target dataset2026.03 | 0.198 | 0.259 | |
| CSTAModality=Visual2025.04 | 0.194 | 0.255 | |
| CSTA2026.03 | 0.194 | 0.255 | |
| CSTA2026.03 | 0.194 | 0.255 | |
| MSVAModality=Visual2025.04 | 0.19 | 0.21 | |
| SSPVS+Textsupervision=supervised, input_modality=video+text2022.01 | 0.181 | 0.238 | |
| SSPVSModality=Visual + Text2025.04 | 0.181 | 0.238 | |
| SSPVS2026.03 | 0.181 | 0.238 | |
| SSPVS2026.03 | 0.181 | 0.238 | |
| UMT2026.03 | 0.179 | 0.235 | |
| UMT2026.03 | 0.179 | 0.235 | |
| VASTSumSplit Protocol=TVT, Summary Budget=15%2026.05 | 0.179 | 0.27 | |
| Human2021.09 | 0.177 | 0.204 | |
| Human2021.07 | 0.177 | 0.204 | |
| HumanSupervision=Unsupervised/Weakly Supervised2021.12 | 0.177 | 0.204 | |
| SSPVSsupervision=supervised, input_modality=video2022.01 | 0.177 | 0.233 | |
| Human2025.04 | 0.177 | 0.204 | |
| HumanSplit Protocol=TVT, Summary Budget=15%2026.05 | 0.177 | 0.204 | |
| SummDiffSplit Protocol=TVT, Summary Budget=15%2026.05 | 0.173 | 0.226 | |
| SSPVS2026.03 | 0.171 | 0.226 | |
| SSPVS2026.03 | 0.171 | 0.226 | |
| CSTA2026.03 | 0.168 | 0.221 | |
| CSTA2026.03 | 0.168 | 0.221 | |
| CSTASplit Protocol=TVT, Summary Budget=15%2026.05 | 0.168 | 0.221 | |
| Joint-VA2026.03 | 0.166 | 0.22 | |
| Joint-VA2026.03 | 0.166 | 0.22 | |
| VASNetModality=Visual2025.04 | 0.16 | 0.17 | |
| VASNet2026.03 | 0.16 | 0.17 | |
| VASNet2026.03 | 0.16 | 0.17 | |
| A2Summ2026.03 | 0.157 | 0.206 | |
| PGL-SUM2026.03 | 0.157 | 0.206 | |
| A2Summ2026.03 | 0.157 | 0.206 | |
| PGL-SUM2026.03 | 0.157 | 0.206 | |
| A2SummSplit Protocol=TVT, Summary Budget=15%2026.05 | 0.157 | 0.206 | |
| Argaw et al.Modality=Visual + Text2025.04 | 0.155 | 0.186 | |
| VASNet2026.03 | 0.153 | 0.205 | |
| VASNet2026.03 | 0.153 | 0.205 | |
| VASNetSplit Protocol=TVT, Summary Budget=15%2026.05 | 0.153 | 0.205 | |
| UMT2026.03 | 0.144 | 0.189 | |
| UMT2026.03 | 0.144 | 0.189 | |
| Joint-VA2026.03 | 0.142 | 0.188 | |
| Joint-VA2026.03 | 0.142 | 0.188 | |
| PGL-SUM2026.03 | 0.141 | 0.186 | |
| PGL-SUM2026.03 | 0.141 | 0.186 | |
| PGL-SUMSplit Protocol=TVT, Summary Budget=15%2026.05 | 0.141 | 0.186 | |
| A2SummModality=Visual + Text2025.04 | 0.137 | 0.165 | |
| A2Summ2026.03 | 0.137 | 0.165 | |
| A2Summ2026.03 | 0.137 | 0.165 | |
| iPTNetModality=Visual2025.04 | 0.134 | 0.163 | |
| DSNet-AFModality=Visual2025.04 | 0.113 | 0.138 | |
| CLIP-It2021.07 | 0.108 | 0.147 | |
| DSNet-ABModality=Visual2025.04 | 0.108 | 0.129 | |
| CLIP-ItModality=Visual + Text2025.04 | 0.108 | 0.147 | |
| VJMHTSupervision=Supervised2021.12 | 0.097 | 0.105 | |
| HMT2021.09 | 0.096 | 0.107 | |
| Park et al. (SumGraph)2021.07 | 0.094 | 0.138 | |
| SumGraphSupervision=Supervised2021.12 | 0.094 | 0.138 | |
| SumGraphsupervision=unsupervised2022.01 | 0.094 | 0.138 | |
| RSGNSupervision=Supervised2021.12 | 0.083 | 0.09 | |
| RSGNsupervision=supervised2022.01 | 0.083 | 0.09 | |
| HSA-RNN2021.09 | 0.082 | 0.088 | |
| VASNet2021.09 | 0.082 | 0.088 | |
| HSA-RNNSupervision=Supervised2021.12 | 0.082 | 0.088 | |
| HSAsupervision=supervised2022.01 | 0.082 | 0.088 | |
| Transformer2021.09 | 0.081 | 0.095 | |
| WS-HRL2021.09 | 0.078 | 0.116 | |
| WS-HRLSupervision=Unsupervised/Weakly Supervised2021.12 | 0.078 | 0.116 | |
| WS-HRLsupervision=unsupervised2022.01 | 0.078 | 0.116 | |
| dppLSTM-att2021.09 | 0.076 | 0.081 | |
| vsLSTM-att2021.09 | 0.074 | 0.083 | |
| VJMHTSupervision=Unsupervised/Weakly Supervised2021.12 | 0.07 | 0.075 | |
| GLRPEsupervision=unsupervised2022.01 | 0.07 | 0.091 | |
| LLMModality=Visual + Text2025.04 | 0.051 | 0.056 | |
| RSGNSupervision=Unsupervised/Weakly Supervised2021.12 | 0.048 | 0.052 | |
| RSGN_usupervision=unsupervised2022.01 | 0.048 | 0.052 | |
| dppLSTM2021.09 | 0.042 | 0.055 | |
| Zhang et al.2021.07 | 0.042 | 0.055 | |
| dppLSTMsupervision=unsupervised2022.01 | 0.042 | 0.055 | |
| CSNetssupervision=unsupervised2022.01 | 0.025 | 0.034 | |
| SUM-GANSupervision=Unsupervised/Weakly Supervised2021.12 | 0.024 | 0.032 | |
| SGANsupervision=unsupervised2022.01 | 0.024 | 0.032 | |
| DR-DSN2021.09 | 0.02 | 0.026 | |
| Zhou et al.2021.07 | 0.02 | 0.026 | |
| DR-DSNSupervision=Unsupervised/Weakly Supervised2021.12 | 0.02 | 0.026 | |
| DRDSNsupervision=unsupervised2022.01 | 0.02 | 0.026 | |
| Random selection2021.09 | 0 | 0 | |
| RandomSupervision=Unsupervised/Weakly Supervised2021.12 | 0 | 0 | |
| Random2025.04 | 0 | 0 | |
| RandomSplit Protocol=TVT, Summary Budget=15%2026.05 | 0 | 0 |