为了分析ICE (Intelligent Communication Environment:智能陆空通话自主训练平台)软件计算机自动评分的有效性,邀请8位评分员分别为2021-2023学年736名学生的27227条陆空通话口语考试数据评分,利用Pearson相关性系数、人机评分一致性...为了分析ICE (Intelligent Communication Environment:智能陆空通话自主训练平台)软件计算机自动评分的有效性,邀请8位评分员分别为2021-2023学年736名学生的27227条陆空通话口语考试数据评分,利用Pearson相关性系数、人机评分一致性系数和等级一致率对比分析机评分和人评分。结果表明:机器评分比人工评分稍低,尤其是汉译英题目;人、机评分相关性系数、一致性系数和一致率均较好。利用Many-Facet Rasch模型分析可知人工评分员内在的一致性和稳定性较好,但评分的严厉度还是存在显著差异。展开更多
This paper proposes a maximum a posteriori （MAP） based blocking artifact reduction algorithm for discrete cosine transform （DCT） domain distributed video coding, in which the SI and the initial reconstructed Wyner...This paper proposes a maximum a posteriori （MAP） based blocking artifact reduction algorithm for discrete cosine transform （DCT） domain distributed video coding, in which the SI and the initial reconstructed Wyner-Ziv （WZ） frame are utilized to further estimate the original WZ frame. Though the MAP estimate improves quality of the artifact region, it also leads to over-smoothness and decreases quality of the non-artifact region. To overcome this problem, a criterion is presented to discriminate the artifact and the non-artifact region in the initial reconstructed WZ frame, and only the artifact region is updated with the MAP estimate. Simulation results show that the proposed algorithm provides obvious improvement in terms of both objective and subjective evaluations.展开更多
Side information （SI） is one of the key issues in distributed video coding （DVC） and affects the compression performance of DVC largely. This paper proposes an SI refinement algorithm, in which the Wyner-Ziv （WZ...Side information （SI） is one of the key issues in distributed video coding （DVC） and affects the compression performance of DVC largely. This paper proposes an SI refinement algorithm, in which the Wyner-Ziv （WZ） frame is split into two parts based on checkerboard pattern, and the two parts are encoded independently but decoded sequentially. In the decoding process, the part 1 is first decoded with the initial SI and partially decoded part （PDP） 1 is used to improve the motion vectors （MVs） and SI of both parts. At the next stage, the part 2 is decoded with the improved SI and PDP 2 is used to further refine MVs of the part 2. Then, SI of both parts are further refined. Simulation results show that the proposed algorithm can improve the peak signal to noise ratio （PSNR） by up to 1.43 dB when compared with traditional DVC codec.展开更多
Depth estimation is a fundamental computer vision problem that infers three-dimensional(3D)structures from a given scene.As it is an ill-posed problem,to fit the projection function from the given scene to the 3D stru...Depth estimation is a fundamental computer vision problem that infers three-dimensional(3D)structures from a given scene.As it is an ill-posed problem,to fit the projection function from the given scene to the 3D structure,traditional methods generally require mass amounts of annotated data.Such pixel-level annotation is quite labor consuming,especially when addressing reflective surfaces such as mirrors or water.The widespread application of deep learning further intensifies the demand for large amounts of annotated data.Therefore,it is urgent and necessary to propose a framework that is able to reduce the requirement on the amount of data.In this paper,we propose a novel semisupervised learning framework to infer the 3D structure from the given scene.First,semantic information is employed to make the depth inference more accurate.Second,we make both the depth estimation and semantic segmentation coarse-to-fine frameworks;thus,the depth estimation can be gradually guided by semantic segmentation.We compare our model with state-of-the-art methods.The experimental results demonstrate that our method is better than many supervised learning-based methods,which proves the effectiveness of the proposed method.展开更多
文摘为了分析ICE (Intelligent Communication Environment:智能陆空通话自主训练平台)软件计算机自动评分的有效性,邀请8位评分员分别为2021-2023学年736名学生的27227条陆空通话口语考试数据评分,利用Pearson相关性系数、人机评分一致性系数和等级一致率对比分析机评分和人评分。结果表明:机器评分比人工评分稍低,尤其是汉译英题目;人、机评分相关性系数、一致性系数和一致率均较好。利用Many-Facet Rasch模型分析可知人工评分员内在的一致性和稳定性较好,但评分的严厉度还是存在显著差异。
基金Supported by the National Natural Science Foundation of China （No.60672088, No.60736043） the National Basic Research Development Program of China （2009CB320905）
文摘This paper proposes a maximum a posteriori （MAP） based blocking artifact reduction algorithm for discrete cosine transform （DCT） domain distributed video coding, in which the SI and the initial reconstructed Wyner-Ziv （WZ） frame are utilized to further estimate the original WZ frame. Though the MAP estimate improves quality of the artifact region, it also leads to over-smoothness and decreases quality of the non-artifact region. To overcome this problem, a criterion is presented to discriminate the artifact and the non-artifact region in the initial reconstructed WZ frame, and only the artifact region is updated with the MAP estimate. Simulation results show that the proposed algorithm provides obvious improvement in terms of both objective and subjective evaluations.
基金Supported by the National Natural Science Foundation of China （ No. 60736043, 60672088） and the National Basic Research Program of China （No. 2009CB32005）.
文摘Side information （SI） is one of the key issues in distributed video coding （DVC） and affects the compression performance of DVC largely. This paper proposes an SI refinement algorithm, in which the Wyner-Ziv （WZ） frame is split into two parts based on checkerboard pattern, and the two parts are encoded independently but decoded sequentially. In the decoding process, the part 1 is first decoded with the initial SI and partially decoded part （PDP） 1 is used to improve the motion vectors （MVs） and SI of both parts. At the next stage, the part 2 is decoded with the improved SI and PDP 2 is used to further refine MVs of the part 2. Then, SI of both parts are further refined. Simulation results show that the proposed algorithm can improve the peak signal to noise ratio （PSNR） by up to 1.43 dB when compared with traditional DVC codec.
基金supported in part by the National High Technology Research and Development Program of China(Grant No.2021YFF0900500)the National Natural Science Foundation of China(Grant Nos.61972115 and 61872116)。
文摘Depth estimation is a fundamental computer vision problem that infers three-dimensional(3D)structures from a given scene.As it is an ill-posed problem,to fit the projection function from the given scene to the 3D structure,traditional methods generally require mass amounts of annotated data.Such pixel-level annotation is quite labor consuming,especially when addressing reflective surfaces such as mirrors or water.The widespread application of deep learning further intensifies the demand for large amounts of annotated data.Therefore,it is urgent and necessary to propose a framework that is able to reduce the requirement on the amount of data.In this paper,we propose a novel semisupervised learning framework to infer the 3D structure from the given scene.First,semantic information is employed to make the depth inference more accurate.Second,we make both the depth estimation and semantic segmentation coarse-to-fine frameworks;thus,the depth estimation can be gradually guided by semantic segmentation.We compare our model with state-of-the-art methods.The experimental results demonstrate that our method is better than many supervised learning-based methods,which proves the effectiveness of the proposed method.
