Ashish Patel 🇮🇳’s Post

𝗗𝗮𝘆-𝟮𝟬𝟯 𝗖𝗼𝗺𝗽𝘂𝘁𝗲𝗿 𝗩𝗶𝘀𝗶𝗼𝗻 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴𝗩𝗶𝗱𝗲𝗼𝗠𝗼𝗖𝗼: Contrastive Video Representation Learning with Temporally Adversarial Examples by Tencent Follow me for a similar post:  🇮🇳 Ashish Patel Interesting Facts : 🔸 This is a paper in #CVPR2020 with over 3 citations. 🔸 It outperforms O3N, Spatio-Temp, VCOP etc. ------------------------------------------------------------------- 𝗔𝗺𝗮𝘇𝗶𝗻𝗴 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 : https://lnkd.in/e92eb5Y Code : https://lnkd.in/eKEK59z ------------------------------------------------------------------- 𝗜𝗠𝗣𝗢𝗥𝗧𝗔𝗡𝗖𝗘 🔸 MoCo is effective for unsupervised image representation learning. In this paper, we propose VideoMoCo for unsupervised video representation learning. Given a video sequence as an input sample, we improve the temporal feature representations of MoCo from two perspectives.  🔸  First, we introduce a generator to drop out several frames from this sample temporally. The discriminator is then learned to encode similar feature representations regardless of frame removals. By adaptively dropping out different frames during training iterations of adversarial learning, we augment this input sample to train a temporally robust encoder. Second, we use temporal decay to model key attenuation in the memory queue when computing the contrastive loss.  🔸 As the momentum encoder updates after keys enqueue, the representation ability of these keys degrades when we use the current input sample for contrastive learning. This degradation is reflected via temporal decay to attend the input sample to recent keys in the queue.  🔸 As a result, we adapt MoCo to learn video representations without empirically designing pretext tasks. By empowering the temporal robustness of the encoder and modelling the temporal decay of the keys, our VideoMoCo improves MoCo temporally based on contrastive learning. Experiments on benchmark datasets including UCF101 and HMDB51 show that VideoMoCo stands as a state-of-the-art video representation learning method. #computervision #artificialintelligence #deeplearning