Object-centric Representation Learning From Unlabeled Videos

Abstract

Supervised (pre-)training currently yields state-of-the-art performance for representation learning for visual recognition, yet it comes at the cost of (1) intensive manual annotations and (2) an inherent restriction in the scope of data relevant for learning. In this work, we explore unsupervised feature learning from unlabeled video. We introduce a novel object-centric approach to temporal coherence that encourages similar representations to be learned for object-like regions segmented from nearby frames. Our framework relies on a Siamese-triplet network to train a deep convolutional neural network (CNN) representation. Compared to existing temporal coherence methods, our idea has the advantage of lightweight preprocessing of the unlabeled video (no tracking required) while still being able to extract object-level regions from which to learn invariances. Furthermore, as we show in results on several standard datasets, our method typically achieves substantial accuracy gains over comp

Stats

• citations8
• S2 citations—
• github stars0
• HF likes0
• heat score7.16
• arxiv keygao2016object

Related papers

• Representation Learning Via Global Temporal Alignment And Cycle-consistency (2021)12.81
• Helping Hands: An Object-aware Ego-centric Video Recognition Model (2023)10.07
• TCLR: Temporal Contrastive Learning For Video Representation (2021)15.78
• Video-text Pre-training With Learned Regions (2021)0.00
• Self-supervised Video Representation Learning Using Inter-intra Contrastive Framework (2020)18.58
• Revisiting Contrastive Methods For Unsupervised Learning Of Visual Representations (2021)3.91
• Multimodal Contrastive Training For Visual Representation Learning (2021)16.32
• Cycle-contrast For Self-supervised Video Representation Learning (2020)0.00