Search results
Apr 23, 2020 · This work uses new features: WRC, CSP, CmBN, SAT, Mish activation, Mosaic data augmentation, C mBN, DropBlock regularization, and CIoU loss, and combine some of them to achieve state-of-the-art results: 43.5% AP for the MS COCO dataset at a realtime speed of ~65 FPS on Tesla V100. There are a huge number of features which are said to improve Convolutional Neural Network (CNN) accuracy ...
May 11, 2023 · This paper conducts a systematic and comprehensive study on vision-language instruction tuning based on the pretrained BLIP-2 models, and introduces an instruction-aware Query Transformer, which extracts informative features tailored to the given instruction. Large-scale pre-training and instruction tuning have been successful at creating general-purpose language models with broad competence ...
Jan 24, 2023 · A statistical test for detecting the watermark with interpretable p-values is proposed, and an information-theoretic framework for analyzing the sensitivity of the watermarks is derived. Potential harms of large language models can be mitigated by watermarking model output, i.e., embedding signals into generated text that are invisible to humans but algorithmically detectable from a short span ...
Oct 10, 2023 · The iTransformer model achieves state-of-the-art on challenging real-world datasets, which further empowers the Transformer family with promoted performance, generalization ability across different variates, and better utilization of arbitrary lookback windows, making it a nice alternative as the fundamental backbone of time series forecasting. The recent boom of linear forecasting models ...
Nov 14, 2014 · The key insight is to build “fully convolutional” networks that take input of arbitrary size and produce correspondingly-sized output with efficient inference and learning. Convolutional networks are powerful visual models that yield hierarchies of features. We show that convolutional networks by themselves, trained end-to-end, pixels-to-pixels, exceed the state-of-the-art in semantic ...
Jul 1, 2002 · A framework for high quality splatting based on elliptical Gaussian kernels is presented, and it is shown that EWA volume reconstruction kernels can be reduced to surface reconstruction kernels, which makes the splat primitive universal in rendering surface and volume data. We present a framework for high quality splatting based on elliptical Gaussian kernels. To avoid aliasing artifacts, we ...
Feb 7, 2018 · This work extends DeepLabv3 by adding a simple yet effective decoder module to refine the segmentation results especially along object boundaries and applies the depthwise separable convolution to both Atrous Spatial Pyramid Pooling and decoder modules, resulting in a faster and stronger encoder-decoder network. Spatial pyramid pooling module or encode-decoder structure are used in deep neural ...
