Startseite Neuigkeiten Neueste Forschungsarbeiten Tutorials Datensätze Wiki SOTA LLM-Modelle GPU-Rangliste Veranstaltungen

Deutsch

Action Recognition In Videos On Activitynet

Metriken

mAP

Ergebnisse

Leistungsergebnisse verschiedener Modelle zu diesem Benchmark

Modellname	mAP	Paper Title	Repository
Ada3D	84.0	2D or not 2D? Adaptive 3D Convolution Selection for Efficient Video Recognition	-
VGG19 + 393K webcam images	53.8	Do Less and Achieve More: Training CNNs for Action Recognition Utilizing Action Images from the Web	-
Text4Vis (w/ ViT-L)	96.9	Revisiting Classifier: Transferring Vision-Language Models for Video Recognition
InternVideo2-6B	95.9	InternVideo2: Scaling Foundation Models for Multimodal Video Understanding
VGG19	52.3	Do Less and Achieve More: Training CNNs for Action Recognition Utilizing Action Images from the Web	-
SMART	84.4	SMART Frame Selection for Action Recognition	-
CD-UAR	53.8	Towards Universal Representation for Unseen Action Recognition	-
P3D	78.9	Learning Spatio-Temporal Representation with Pseudo-3D Residual Networks
RRA	83.4	Fine-grained Video Categorization with Redundancy Reduction Attention	-
MARL (w/ SEResNeXt-152)	90.05	Multi-Agent Reinforcement Learning Based Frame Sampling for Effective Untrimmed Video Recognition	-
DSN	87.9	Dynamic Sampling Networks for Efficient Action Recognition in Videos	-
NSNet (w/ Swin-L)	94.3	NSNet: Non-saliency Suppression Sampler for Efficient Video Recognition	-
ListenToLook	89.9	Listen to Look: Action Recognition by Previewing Audio
DSANet (w/ 3D ResNet50)	90.5	DSANet: Dynamic Segment Aggregation Network for Video-Level Representation Learning
BIKE	96.1	Bidirectional Cross-Modal Knowledge Exploration for Video Recognition with Pre-trained Vision-Language Models
TSQNet (w/ Swin-L)	93.7	Temporal Saliency Query Network for Efficient Video Recognition	-

0 of 16 row(s) selected.