第 1 章 緒論 ····························································································1
1.1 邊云智能產(chǎn)生的大背景····································································1
1.1.1 新一代信息技術(shù)的快速發(fā)展·····················································2
1.1.2 國(guó)家政策的支持與引導(dǎo)···························································6
1.2 邊云智能······················································································7
1.3 邊云智能的發(fā)展·············································································9
1.3.1 邊云智能的三大發(fā)展階段························································9
1.3.2 城市大腦··········································································.11
1.4 “智能+”新潮頭··········································································.13
1.4.1 “智能+”技術(shù)新融合···························································.13
1.4.2 多維度場(chǎng)景應(yīng)用·································································.14
本章習(xí)題··························································································.15
第 2 章 邊云架構(gòu) ···················································································.16
2.1 系統(tǒng)工程方法論··········································································.17
2.1.1 概述 ················································································.17
2.1.2 基本方法··········································································.17
2.2 邊云智能體系架構(gòu)模型·································································.20
2.2.1 概念框架··········································································.20
2.2.2 層次結(jié)構(gòu)··········································································.22
2.3 協(xié)同模式···················································································.23
2.3.1 “云-邊”協(xié)同 ····································································.24
2.3.2 “邊-邊”協(xié)同 ····································································.25
2.3.3 “邊-端”協(xié)同 ····································································.27
2.3.4 “云-邊-端”協(xié)同 ································································.28
2.3.5 度量指標(biāo)··········································································.28
2.4 邊云智能架構(gòu)應(yīng)用·······································································.30
2.4.1 “云-邊-端”區(qū)塊鏈 ·····························································.30
2.4.2 “云-邊-端”一體化機(jī)器人系統(tǒng) ··············································.32
本章習(xí)題··························································································.33
第 3 章 深度學(xué)習(xí) ···················································································.35
3.1 深度學(xué)習(xí)概念·············································································.36
3.1.1 人工智能與機(jī)器學(xué)習(xí)···························································.36
3.1.2 深度學(xué)習(xí)··········································································.37
3.1.3 神經(jīng)網(wǎng)絡(luò)··········································································.39
3.2 前饋神經(jīng)網(wǎng)絡(luò)·············································································.39
3.2.1 感知機(jī)模型·······································································.39
3.2.2 反向傳播··········································································.42
3.2.3 卷積神經(jīng)網(wǎng)絡(luò)····································································.44
3.2.4 幾種典型的卷積神經(jīng)網(wǎng)絡(luò)·····················································.47
3.3 反饋神經(jīng)網(wǎng)絡(luò)·············································································.50
3.3.1 循環(huán)神經(jīng)網(wǎng)絡(luò)····································································.50
3.3.2 長(zhǎng)短期神經(jīng)網(wǎng)絡(luò)·································································.53
3.4 Transformer 神經(jīng)網(wǎng)絡(luò) ···································································.56
3.4.1 編碼器單元與解碼器單元·····················································.58
3.4.2 多頭注意力機(jī)制·································································.59
3.4.3 非參位置編碼····································································.60
本章習(xí)題··························································································.61
第 4 章 自然語(yǔ)言處理 ·············································································.62
4.1 自然語(yǔ)言處理概述·······································································.63
4.1.1 自然語(yǔ)言處理簡(jiǎn)介······························································.63
4.1.2 自然語(yǔ)言處理的發(fā)展歷史·····················································.74
4.1.3 自然語(yǔ)言處理的應(yīng)用及面臨的挑戰(zhàn)·········································.76
4.2 文本挖掘···················································································.79
4.2.1 文本挖掘簡(jiǎn)介····································································.79
4.2.2 文本挖掘算法····································································.81
4.3 機(jī)器翻譯···················································································.87
4.3.1 機(jī)器翻譯簡(jiǎn)介····································································.87
4.3.2 機(jī)器翻譯算法····································································.89
4.4 自動(dòng)問(wèn)答系統(tǒng)·············································································.93
4.4.1 自動(dòng)問(wèn)答系統(tǒng)簡(jiǎn)介······························································.93
4.4.2 自動(dòng)問(wèn)答系統(tǒng)模型······························································.95
4.5 語(yǔ)音識(shí)別···················································································101
4.5.1 語(yǔ)音識(shí)別簡(jiǎn)介····································································102
4.5.2 語(yǔ)音識(shí)別算法····································································103
本章習(xí)題··························································································105
第 5 章 計(jì)算機(jī)視覺(jué) ················································································107
5.1 計(jì)算機(jī)視覺(jué)概述··········································································107
5.1.1 計(jì)算機(jī)視覺(jué)簡(jiǎn)介·································································108
5.1.2 計(jì)算機(jī)視覺(jué)的發(fā)展歷史························································109
5.1.3 計(jì)算機(jī)視覺(jué)的應(yīng)用及面臨的挑戰(zhàn)···········································.110
5.2 圖像分類(lèi)··················································································.114
5.2.1 圖像分類(lèi)簡(jiǎn)介···································································.114
5.2.2 圖像分類(lèi)算法···································································.115
5.3 目標(biāo)檢測(cè)··················································································.119
5.3.1 目標(biāo)檢測(cè)簡(jiǎn)介···································································.119
5.3.2 目標(biāo)檢測(cè)算法····································································120
5.4 圖像分割···················································································123
5.4.1 圖像分割簡(jiǎn)介····································································123
5.4.2 圖像分割算法····································································124
5.5 目標(biāo)跟蹤···················································································125
5.5.1 目標(biāo)跟蹤簡(jiǎn)介····································································126
5.5.2 目標(biāo)跟蹤算法····································································126
本章習(xí)題··························································································128
第 6 章 邊緣輕量化 ················································································129
6.1 邊緣輕量化的簡(jiǎn)介·······································································129
6.1.1 邊緣端對(duì)輕量化的需求························································129
6.1.2 什么是邊緣輕量化······························································130
6.2 模型壓縮方法·············································································131
6.2.1 量化和二值化····································································131
6.2.2 網(wǎng)絡(luò)剪枝··········································································131
6.2.3 低秩因子分解····································································132
6.2.4 參數(shù)共享··········································································133
6.2.5 蒸餾學(xué)習(xí)··········································································133
6.2.6 加速網(wǎng)絡(luò)設(shè)計(jì)····································································134
6.3 模型壓縮舉例·············································································137
6.3.1 知識(shí)蒸餾··········································································137
6.3.2 深度壓縮··········································································139
6.3.3 MNASNet ·········································································143
本章習(xí)題··························································································145
第 7 章 云端決策 ···················································································146
7.1 云端決策簡(jiǎn)介·············································································147
7.1.1 云端決策的重要性······························································147
7.1.2 云端決策的特點(diǎn)·································································147
7.2 云端決策——大數(shù)據(jù)挖掘······························································149
7.2.1 回歸分析··········································································149
7.2.2 聚類(lèi) ················································································150
7.2.3 關(guān)聯(lián)規(guī)則··········································································152
7.3 云端決策——推薦算法·································································154
7.3.1 基于統(tǒng)計(jì)的推薦算法···························································155
7.3.2 基于協(xié)同過(guò)濾的推薦系統(tǒng)·····················································155
7.3.3 基于內(nèi)容的推薦系統(tǒng)···························································156
7.3.4 基于關(guān)聯(lián)規(guī)則的推薦系統(tǒng)·····················································158
7.3.5 基于網(wǎng)絡(luò)結(jié)構(gòu)的推薦系統(tǒng)·····················································158
本章習(xí)題··························································································159
第 8 章 邊云智能賦能智慧教室 ·································································160
8.1 智慧教室的形成背景與邊云框架·····················································161
8.1.1 智慧教室政策支持與特征分析 ··············································162
8.1.2 基于邊云智能的智慧教室框架 ··············································164
8.1.3 基于邊云智能建設(shè)的智慧教室目標(biāo)愿景 ··································166
8.2 智慧教室的邊緣端感知技術(shù)與應(yīng)用··················································166
8.2.1 無(wú)感考勤、表情感知與異常行為識(shí)別 ·····································167
8.2.2 邊緣端感知模型的壓縮與輕量化 ···········································172
8.3 智慧教室的云端決策技術(shù)與應(yīng)用·····················································174
8.3.1 “教育大腦”大數(shù)據(jù)分析決策方法··········································174
8.3.2 個(gè)性化推薦、學(xué)習(xí)評(píng)價(jià)與師生互動(dòng)應(yīng)用 ··································176
8.4 “邊云智能+”前景展望·································································178
8.4.1 邊云智能賦能智慧交通························································178
8.4.2 邊云智能賦能智慧安防························································185
本章習(xí)題··························································································190
習(xí)題答案································································································191