Technology peripheralsAIYour background is too fake. It was automatically synthesized using AI. Is it fake?
Hello, everyone.
Have you all seen the video below a few days ago?
The blogger originally wanted to prove that his background was real, but as a result, netizens made fun of him and changed into various backgrounds to "fight the fake".
Let’s join in the fun today and use AI technology to automatically replace the background.
The idea is not difficult. We first separate the characters from the original video, and then "paste" the separated characters into the new background video.
The key technology used to classify people from videos is instance segmentation in computer vision. I have introduced target detection technology in my previous article, such as face detection
Target detection is to mark the detected target through a rectangular box, which is relatively easy. Instance segmentation is a more delicate task because each pixel needs to be classified and the outline of the object is accurately outlined.
We use detectron2 for instance segmentation today. It is an open source project of Facebook AI Research Institute. It is powerful and easy to use.
1. Installation
detectron2 supports installation on Linux and macOS systems. Linux can be installed directly through pip, while mac can only To install through source code compilation, it is recommended that you use Linux.
Supports GPU and CPU running. I am using a 3090 graphics card, CUDA11.1, and Pytorch1.8.
2. Run
We use the official pre-trained model to perform instance segmentation on the image.
2.1 Load the model configuration file
<span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">from</span> detectron2<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.config</span> import get_cfg<br><span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">from</span> detectron2 import model_zoo<br><br>cfg <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> get_cfg<span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">(</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">)</span><br>model_cfg_file <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> model_zoo<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.get_config_file</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">(</span><span style="color: rgb(102, 153, 0); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">'COCO-InstanceSegmentation/mask_rcnn_R_50_FPN_3x.yaml'</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">)</span><br>cfg<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.merge_from_file</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">(</span>model_cfg_file<span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">)</span>
COCO-InstanceSegmentationRepresents the instance segmentation model trained with the cocodata set.
mask_rcnn_R_50_FPN_3x.yaml is the configuration information used for model training.
As you can see from the picture below, in addition to providing instance segmentation models, detectron2 also provides target detection, key point detection and other models, which is quite comprehensive.
2.2 Loading the model
model_weight_url <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> model_zoo<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.get_checkpoint_url</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">(</span><span style="color: rgb(102, 153, 0); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">'COCO-InstanceSegmentation/mask_rcnn_R_50_FPN_3x.yaml'</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">)</span><br>cfg<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.MODEL</span><span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.WEIGHTS</span> <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> model_weight_url
mask_rcnn_R_50_FPN_3x.yaml file stores the url of the pre-trained model. When performing instance segmentation, the program will automatically download the model from the URL to the local. The storage location is:
# However, the automatic download method of the program may be slower. At this time, you can use Thunder to download the model file yourself and put it in the corresponding path. .
2.3 Instance segmentation
First, read a picture, the picture size is 480 * 640.
img <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> cv2<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.imread</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">(</span><span style="color: rgb(102, 153, 0); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">'./000000439715.jpg'</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">)</span>
Instantiate the DefaultPredictor object.
<span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">from</span> detectron2<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.engine</span> import DefaultPredictor<br>predictor <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> DefaultPredictor<span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">(</span>cfg<span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">)</span>
Instance segmentation of images
out <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> predictor<span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">(</span>img<span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">)</span>
outThe variable stores the category id, detection frame and target mask of each segmented target.
out["instances"].pred_classesGet the category id of the target
A total of 15 targets have been detected here, and the categories can be found in the configuration file Mapping relationship between id and category name. Among them, 0 represents a person and 17 represents a horse.
out["instances"].pred_masksGet the mask of the target. Let's take the mask of a single target to study its use.
As you can see, its value is of Boolean type, and the shape is the same size as the image.
所以，遮罩是实例分割的结果，里面每个元素对应Your background is too fake. It was automatically synthesized using AI. Is it fake?一个像素，取值为True代表该像素是检测出来的目标像素。
因此，我们可以通过遮罩给目标加上一层不透明度，从而把目标精确标注出来。
<span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">img_copy</span> <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> <span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">img</span><span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.</span><span style="color: rgb(153, 0, 85); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">copy</span>()<br><br><span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">alpha</span> <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> <span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">0</span><span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.8</span><br><span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">color</span> <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> <span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">np</span><span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.</span><span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">array</span>([<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">0</span>, <span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">255</span>, <span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">0</span>])<br><br><span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">img_copy</span>[<span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">mask</span> <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">></span> <span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">0</span>, :] <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> <span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">img_copy</span>[<span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">mask</span> <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">></span> <span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">0</span>, :] <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">*</span> <span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">alpha</span> <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">+</span> <span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">color</span> <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">*</span> (<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">1</span><span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">-</span><span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">alpha</span>)
上述给目标加上一层绿色的不透明度，效果如下：
可以看到，骑在马上的人已经被标注出来了。
3. 自动合成背景
有了上面的基础，我们就很容易合成视频了。
读取原视频每一帧中将人物分割出来，将分割出来的人物直接覆盖到新背景视频中对应的帧即可。
核心代码：
# 读取原视频<br>ret<span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">,</span> frame_src <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> cap<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.read</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">(</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">)</span><br><br># 读取新背景视频<br>ret<span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">,</span> frame_bg <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> cap_bg<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.read</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">(</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">)</span><br># 调整背景尺寸跟原视频一样<br>frame_bg <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> cv2<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.resize</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">(</span>frame_bg<span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">,</span> sz<span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">)</span><br><br># 分割原视频人物<br>person_mask <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> instance_seg<span style="color: rgb(0, 92, 197); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">.predict</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">(</span>frame_src<span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">)</span><br># 合成<br>frame_bg<span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">[</span>person_mask<span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">,</span> <span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">:</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">]</span> <span style="color: rgb(215, 58, 73); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">=</span> frame_src<span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">[</span>person_mask<span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">,</span> <span style="color: rgb(89, 89, 89); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">:</span><span style="color: rgb(153, 153, 119); margin: 0px; padding: 0px; background: none 0% 0% / auto repeat scroll padding-box border-box rgba(0, 0, 0, 0);">]</span>
The above is the detailed content of Your background is too fake. It was automatically synthesized using AI. Is it fake?. For more information, please follow other related articles on the PHP Chinese website!
ai合并图层的快捷键是什么Jan 07, 2021 am 10:59 AMai合并图层的快捷键是“Ctrl+Shift+E”，它的作用是把目前所有处在显示状态的图层合并，在隐藏状态的图层则不作变动。也可以选中要合并的图层，在菜单栏中依次点击“窗口”-“路径查找器”，点击“合并”按钮。
ai橡皮擦擦不掉东西怎么办Jan 13, 2021 am 10:23 AMai橡皮擦擦不掉东西是因为AI是矢量图软件，用橡皮擦不能擦位图的，其解决办法就是用蒙板工具以及钢笔勾好路径再建立蒙板即可实现擦掉东西。
谷歌超强AI超算碾压英伟达A100！TPU v4性能提升10倍，细节首次公开Apr 07, 2023 pm 02:54 PM虽然谷歌早在2020年，就在自家的数据中心上部署了当时最强的AI芯片——TPU v4。但直到今年的4月4日，谷歌才首次公布了这台AI超算的技术细节。论文地址：https://arxiv.org/abs/2304.01433相比于TPU v3，TPU v4的性能要高出2.1倍，而在整合4096个芯片之后，超算的性能更是提升了10倍。另外，谷歌还声称，自家芯片要比英伟达A100更快、更节能。与A100对打，速度快1.7倍论文中，谷歌表示，对于规模相当的系统，TPU v4可以提供比英伟达A100强1.
ai可以转成psd格式吗Feb 22, 2023 pm 05:56 PMai可以转成psd格式。转换方法：1、打开Adobe Illustrator软件，依次点击顶部菜单栏的“文件”-“打开”，选择所需的ai文件；2、点击右侧功能面板中的“图层”，点击三杠图标，在弹出的选项中选择“释放到图层（顺序）”；3、依次点击顶部菜单栏的“文件”-“导出”-“导出为”；4、在弹出的“导出”对话框中，将“保存类型”设置为“PSD格式”，点击“导出”即可；
ai顶部属性栏不见了怎么办Feb 22, 2023 pm 05:27 PMai顶部属性栏不见了的解决办法：1、开启Ai新建画布，进入绘图页面；2、在Ai顶部菜单栏中点击“窗口”；3、在系统弹出的窗口菜单页面中点击“控制”，然后开启“控制”窗口即可显示出属性栏。
GPT-4的研究路径没有前途？Yann LeCun给自回归判了死刑Apr 04, 2023 am 11:55 AMYann LeCun 这个观点的确有些大胆。 「从现在起 5 年内，没有哪个头脑正常的人会使用自回归模型。」最近，图灵奖得主 Yann LeCun 给一场辩论做了个特别的开场。而他口中的自回归，正是当前爆红的 GPT 家族模型所依赖的学习范式。当然，被 Yann LeCun 指出问题的不只是自回归模型。在他看来，当前整个的机器学习领域都面临巨大挑战。这场辩论的主题为「Do large language models need sensory grounding for meaning and u
强化学习再登Nature封面，自动驾驶安全验证新范式大幅减少测试里程Mar 31, 2023 pm 10:38 PM引入密集强化学习，用 AI 验证 AI。 自动驾驶汽车 (AV) 技术的快速发展，使得我们正处于交通革命的风口浪尖，其规模是自一个世纪前汽车问世以来从未见过的。自动驾驶技术具有显着提高交通安全性、机动性和可持续性的潜力，因此引起了工业界、政府机构、专业组织和学术机构的共同关注。过去 20 年里，自动驾驶汽车的发展取得了长足的进步，尤其是随着深度学习的出现更是如此。到 2015 年，开始有公司宣布他们将在 2020 之前量产 AV。不过到目前为止，并且没有 level 4 级别的 AV 可以在市场
ai移动不了东西了怎么办Mar 07, 2023 am 10:03 AMai移动不了东西的解决办法：1、打开ai软件，打开空白文档；2、选择矩形工具，在文档中绘制矩形；3、点击选择工具，移动文档中的矩形；4、点击图层按钮，弹出图层面板对话框，解锁图层；5、点击选择工具，移动矩形即可。
Hot AI Tools
Undresser.AI Undress
AI-powered app for creating realistic nude photos
AI Clothes Remover
Online AI tool for removing clothes from photos.
Undress AI Tool
Undress images for free
Clothoff.io
AI clothes remover
AI Hentai Generator
Generate AI Hentai for free.
Hot Article
Hot Tools
Safe Exam Browser
Safe Exam Browser is a secure browser environment for taking online exams securely. This software turns any computer into a secure workstation. It controls access to any utility and prevents students from using unauthorized resources.
SAP NetWeaver Server Adapter for Eclipse
Integrate Eclipse with SAP NetWeaver application server.
SublimeText3 Chinese version
Chinese version, very easy to use
DVWA
Damn Vulnerable Web App (DVWA) is a PHP/MySQL web application that is very vulnerable. Its main goals are to be an aid for security professionals to test their skills and tools in a legal environment, to help web developers better understand the process of securing web applications, and to help teachers/students teach/learn in a classroom environment Web application security. The goal of DVWA is to practice some of the most common web vulnerabilities through a simple and straightforward interface, with varying degrees of difficulty. Please note that this software
Dreamweaver Mac version
Visual web development tools
