Ucf101
PulseAugur coverage of Ucf101 — every cluster mentioning Ucf101 across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
Gaussian Video Transformer advances video representation with 2DGS tokenization
Researchers have developed the Gaussian Video Transformer (GVT), a novel framework for video representation that utilizes a feed-forward 2D Gaussian Splatting (2DGS) tokenization scheme. This approach enhances spatial a…
-
New KD method enhances action recognition model compression
Researchers have developed a novel Channel-wise Dynamic Knowledge Distillation (KD) approach called ASCD KD to improve the compression of large action recognition models. This method addresses limitations in existing KD…
-
Codebook Capacity Dominates Video Compression Quality Over Resolution
Researchers have found that codebook capacity, rather than spatial resolution, is the primary factor influencing perceptual quality in hierarchical discrete video compression. An empirical study using MS-VQ-VAE models a…
-
HyperGS advances video representation with fast, generalizable Gaussian predictions
Researchers have developed HyperGS, a novel feedforward approach that directly predicts Gaussian representations for videos in a single pass, eliminating the need for per-video optimization. This method significantly sp…
-
HyperGS advances Gaussian video representation with faster, generalizable predictions
Researchers have developed HyperGS, a novel feedforward approach for video representation using Gaussian Splatting. Unlike previous methods requiring per-video optimization, HyperGS directly predicts Gaussian representa…
-
New MS-VQ-VAE method achieves ultra-low bitrate video compression
Researchers have developed a novel video compression method using a discrete latent representation called MS-VQ-VAE, which operates effectively at ultra-low bitrates below 0.1 bits per pixel. This approach overcomes lim…
-
New Bit-ViP technique enhances visual privacy in images
Researchers have developed a new image obfuscation technique called Bit-ViP, designed to protect visual privacy in images used for computer vision tasks. This method incorporates non-invertible noise generated by chaoti…
-
New Bit-plane Technique Enhances Image Privacy Against Reconstruction Attacks
Researchers have developed Bit-ViP, a novel image obfuscation technique designed to protect visual privacy in computer vision applications. This method uses bit-planes to secure images, making them resistant to reconstr…
-
New MoFore Framework Advances Self-Supervised Video Representation Learning
Researchers have introduced MoFore, a novel framework for self-supervised video representation learning that focuses on forecasting future latent embeddings from distant context clips. Unlike previous methods that relie…
-
New CFE-PPAR method enables privacy-preserving video action recognition with compression
Researchers have introduced CFE-PPAR, a novel method for privacy-preserving action recognition in videos that remains effective even after compression. Unlike previous approaches that suffer significant performance degr…