Non-line-of-sight imaging over 1.43km
參賽者：
劉健江-中國科學技術大學
非視域成像的目標在於通過探測發射雷射三次漫反射後的回波信號，通過算法重構出隱藏在角落後障礙物後的物體，從而實現「隔牆觀物」，本作品在前人工作基礎上，首次實現了公裡級別的非視域成像。Non-line-of-sight (NLOS) imaging aims to reconstruct the hidden objects from the indirect light paths that scatter multiple times, typically using the information encoded in the time-of-flight of scattered photons. Based on previous work, this work is the first time to realize NLOS imaging over 1.43km range.Perovskite light-emitting diodes based on spontaneously formed submicrometre-scale structures
曹雨-西北工業大學（2019年博士畢業於南京工業大學）近年來，金屬滷化物鈣鈦礦發光二極體發展迅速，吸引了人們極大的關注。我們通過一種簡單的低溫溶液法自組裝形成了具有亞微米結構的鈣鈦礦薄膜，實現了高效、高亮度的鈣鈦礦發光二極體的製備。In recent years, metal halide perovskite light emitting diodes (LEDs) have attracted great attention due to their remarkable performance. Here we achieve high-efficiency and high-brightness perovskite LEDs from solution-processed perovskite films that spontaneously form submicrometre-scale structures.Learning Credit Assignment
為了破解深度學習黑箱，建立神經網絡決策與微觀組分之間的橋梁，我們提出一種由稀疏度與均值方差刻畫的隨機連接權重模型，通過平均場方法訓練得到系綜水平上一組子網絡，並在真實據集上獲得了不弱於傳統方法的表現。
Credit assignment problem (CAP) has long been an interesting topic connecting the macroscopic behaviors with microscopic interactions of components in deep neural networks. To solve this problem and reveal the mystery behind the black box of deep learning, we put forward a model with random weights characterized by a spike and slab distribution, which obtains comparable or even better performance than traditional models. An optimal random network ensemble can be achieved after training based on mean-field theory.
前4批展示作品的連結 （註：禮品助力投票時限已過，助TA挑選禮品二維碼已經失效）：第1批作品展示
第2批作品展示
第3批作品展示
第4批作品展示
點擊「閱讀原文」，直達助力投票頁面