Lane detection, the process of identifying lane markings as approximated curves, is widely used for lane departure warning and adaptive cruise control in autonomous vehicles. The popular pipeline that solves it in two stepsfeature extraction plus post-processing, while useful, is too inefficient and flawed in learning the global context and lanes' long and thin structures. To tackle these issues, we propose an end-to-end method that directly outputs parameters of a lane shape model, using a network built with a transformer to learn richer structures and context. The lane shape model is formulated based on road structures and camera pose, providing physical interpretation for parameters of network output. The transformer models nonlocal interactions with a self-attention mechanism to capture slender structures and global context. The proposed method is validated on the TuSimple benchmark and shows state-of-the-art accuracy with the most lightweight model size and fastest speed. Additionally, our method shows excellent adaptability to a challenging self-collected lane detection dataset, showing its powerful deployment potential in real applications. Codes are available at https: //github.com/liuruijin17/LSTR.

No abstract

No abstract

No abstract

No abstract

Video coding schemes designed based on sequential or predictive coding models are vulnerable to the loss of encoded frames at the decoder end. Motivated by this observation, in this thesis we propose two new coding models: robust sequential coding and robust predictive coding. For the Gauss-Markov source with the mean squared error distortion measure, we characterize certain supporting hyperplanes of the rate region of these two coding problems. The proof is divided into three steps: 1) it is shown that each supporting hyperplane of the rate region of Gaussian robust sequential coding admits a max-min lower bound; 2) the corresponding min-max upper bound is shown to be achievable by a robust predictive coding scheme; 3) a saddle point analysis proves that the max-min lower bound coincides with the min-max upper bound.Furthermore, it is shown that the proposed robust predictive coding scheme can be implemented using a successive quantization system. Theoretical and experimental results indicate that this scheme has a desirable "self-recovery" property. Our investigation also reveals an information-theoretic minimax theorem and the associated extremal inequalities.iv

No abstract

We introduce a novel approach for blind and semi-blind watermarking and apply it to images. We derive randomized robust semi-global statistics of images in a suitable transform domain (wavelets in case of images) and quantize them in order to embed the watermark. Quantization is effectively carried out by embedding to the host a computed sequence, which is obtained by solving an optimization problem whose parameters are known to the information hider but unknown to the attacker. An essential emphasis of the proposed method is randomization, which is crucial for security and robustness against arbitrary quality-preserving attacks. We formally show that malicious optimal estimation attacks that are specifically derived for our algorithm are ineffective in practice. Furthermore, we experimentally demonstrate that our watermarking method survives many generic benchmark attacks for a large number of images.