CCTV13 — broadcast
20261102 163000 UTC
230 transcript segments
Google Cloud Speech-to-Text API (Chirp) + Gemini 2.5 Flash Non-Thinking
Data courtesy of The GDELT Project (https://www.gdeltproject.org/), from the Internet Archive TV News Archive.
Machine transcription. Treat it as a searchable index of what was broadcast, not a verbatim quotation record.

[00:00:00] The silence system in the action area has ended. This system was valid from
[00:00:04] October 27th to November 1st. He said that all parties are currently
[00:00:07] working together to continue humanitarian relief operations. Residents remaining in Aleppo
[00:00:11] still need help. So what is the likelihood of this kind of humanitarian ceasefire or
[00:00:15] rescue operation being replicated in other conflict areas?
[00:00:19] Also, why does the Ukrainian side hope that Turkey will facilitate the meeting?
[00:00:23] Is direct communication between the Russian and Ukrainian leaders appropriate under the current situation?
[00:00:27] 间 的 直 接 沟 通 是 否 到 了 一 个 合 适 的 。
[00:00:50] Through this operation, the Russian side successfully completed the evacuation of civilians from the war zone, which to a certain extent improved its international image
[00:00:54] and also helped to rebuild the political
[00:00:58] trust between Russia and Ukraine through humanitarian
[00:01:01] cooperation. Since Aleppo was not the main
[00:01:05] battlefield of the conflict between Russia and Ukraine, this transfer operation did not affect
[00:01:08] the overall combat rhythm of both sides. The Aleppo
[00:01:12] humanitarian ceasefire model has certain substantive value, but its application
[00:01:16] range is limited to non-core war zones and cannot be
[00:01:20] widely promoted in main war zones. Ukraine seeks Turkey to mediate
[00:01:24] a summit of heads of state, which has clear diplomatic considerations.
[00:01:28] Turkey maintains relations with both Russia and Ukraine.
[00:02:23] If it is held hastily,
[00:02:26] it will probably only repeat the past situation of each side saying its own thing.
[00:02:37] This awkward situation makes it difficult to produce substantive diplomatic achievements. The Belgian Royal Palace issued a statement on November 2nd
[00:02:41] saying that Belgian Prime Minister De Croo submitted his resignation to King Philip
[00:02:45] that day due to the failure of federal government budget negotiations. The king delayed his decision on whether
[00:02:49] to accept the resignation and will begin consultations with the leaders of the ruling coalition
[00:02:52] parties on the 3rd. The statement also said that the king
[00:02:56] will meet with De Croo again on the 6th. Belgium is
[00:03:33] You are watching Five Nights News.
[00:03:37] 。
[00:03:40] The truth is out. We are currently in the peak season for e-commerce promotions.
[00:03:44] In the past two days, two pictures have attracted widespread attention from netizens. Let's
[00:03:47] take a look together. First, the first picture is a seller's show from
[00:03:51] a clothing store on an online shopping platform, which is a model picture of clothing products.
[00:03:55] The model with three legs in the picture made many people. The picture is AI generated, right? Yes, it is AI generated. Many friends commented that the money they spent is real, but how many real products can they see?
[00:03:59] 多 人 。 圖片是AI生成的吧,沒錯,就是AI生成的,不少友表示,我花的錢是真的,但是我能看到幾張真的產品呢?
[00:04:23] In the clothing store, has the application of AI
[00:04:25] model pictures become a trend? With the continuous development of artificial intelligence technology,
[00:04:46] First, let's take a look at the report from China Central Television's reporter. Open major e-commerce platforms and search for clothing-related keywords. A large number
[00:04:50] of product links pop up on the page, with the main image showing models in various styles.
[00:04:54] The reporter then narrowed down to a long-sleeved silk shirt
[00:04:58] product sold in an official flagship store.
[00:05:07] AI
[00:05:08] content is AI,
[00:05:12] but the customer service claimed that the product display image was taken by a real person. Another
[00:05:16] woolen suit jacket sold in a clothing store also came with a short video
[00:05:20] display of the same model in the same scene as the main product image.
[00:05:24] However, the reporter observed that the model's voice and mouth movements were not
[00:05:27] always consistent, and the AI was quite obvious. This video was also
[00:05:31] detected by the National Anti-Fraud Center APP as containing AI-generated
[00:05:35] traces. In the product main images that the reporter subsequently checked, some of the
[00:05:39] model display effects were almost indistinguishable from real photos. However, the detection results
[00:05:43] still showed cases of AI-generated traces. Several e-commerce practitioners in the clothing industry
[00:05:47] confirmed that the use of AI model images is quite common in the current
[00:05:50] industry, especially since this year, many merchants have started using the generative
[00:05:54] seller show model.
[00:05:56] 。
[00:05:59] The fastest it can be is three to four days to get feedback. It basically
[00:06:03] takes only a few minutes if it's slow. I can generate
[00:06:07] the entire set of pictures as long as my details are in place. For example, in the past, if we took a set
[00:06:10] of pictures, the normal price was. AI has become modularized, and some can even be purchased for a few thousand dollars, and the operation is very simple.
[00:06:32] The reporter learned that a certain scale of AI has been formed in the current market, including AI for clothing, etc.
[00:06:42] Some technology companies have developed AI model image generation systems that can
[00:06:46] be operated in a streamlined manner. As long as you upload a picture, the website can complete
[00:06:50] the generation of the model image. Upload the model you want, support uploading one to four
[00:06:54] material images, or upload a section of the display like a puzzle. AI
[00:06:57] will intelligently combine the clothing and background to generate a matching image in a unified style. You can freely choose
[00:07:01] the proportion, style, and template, and generate single images or batch images at once, which can quickly
[00:07:05] generate different poses and natural movements. Several e-commerce practitioners in the clothing industry
[00:07:09] stated that some of the current generative seller shows can already achieve
[00:07:13] almost no difference from the effects of previous real-person model studio shoots and post-editing. The product
[00:07:17] return rate will also not significantly fluctuate due to the use of AI model images
[00:07:21] by e-commerce businesses. Some industry insiders believe that as practitioners become more
[00:07:25] proficient in technology, generative seller shows may be able to help customers
[00:07:29] make more accurate consumption decisions.
[00:07:31] For merchants, AI image generation
[00:07:35] is fast and low-cost, but for consumers, does the rampant use of e-commerce AI model
[00:07:39] images constitute false advertising and consumer fraud? Let's
[00:07:42] hear the expert's interpretation. The reporter
[00:07:45] observed that on e-commerce platforms, merchants whose product model display
[00:07:49] images have been detected as containing AI-generated traces will actively
[00:07:53] label the model as AI-generated or similar text. The reporter also noticed
[00:07:57] that multiple platforms had previously issued announcements to regulate the chaos of AI model images
[00:08:01] and crack down on illegal acts such as using AI fake images to deceive consumers
[00:08:05] and clearing original brand merchants. They promoted the active labeling of AI materials.
[00:08:08] They included AI images that did not match the actual product in some key random inspections or complaints.
[00:08:23] AI generated food,
[00:08:28] However, according to the requirements of depth verification, the backend will probably gradually classify them. Analysis believes that legally, it is not forbidden
[00:08:36] to use AI models normally, but the industry needs to clarify two boundaries. One is the display effect of the image.
[00:08:42] AI effects, coupled with some real fabrics and real shots, are generally fine. But if the merchant claims it's a real person's shot, and the AI's version, fabric, and other differences are so great that they affect the consumer's order, then it may constitute. If
[00:09:54] it's AI, it's hard to be that precise. Compared to real photos, they have that subtle touch, the density of the knit, and other aspects that are closer to reality.
[00:10:04] Does a top get stuck? Does it get stuck? What state is it in when walking? Does a pair of pants pile up at the waist when sitting down? It's uncomfortable.
[00:10:13] AI models cannot display these details. Objectively speaking, the current application of AI
[00:10:16] model images in e-commerce clothing does have shortcomings. On the one hand,
[00:10:20] there are problems with the generative model itself, which can easily lead to physical deformities and human body structures.
[00:10:24] Drape, and it's also hard to create a realistic model that looks good in the picture but has a big difference in the actual fit. The probability of repeated and returned items is even higher.
[00:10:46] From a technical perspective, it is difficult for ordinary text-to-image
[00:10:49] AI to fully display the real effect and satisfy the need for
[00:10:55] sensory experience. Of course, binding sample clothing scanning, fabric
[00:10:58] capture, and 3D cloth simulation vertical virtual solutions
[00:11:02] can theoretically achieve high-fidelity display, but this solution is relatively expensive
[00:11:06] and will be higher. It is currently difficult to popularize it to a large number of small
[00:11:10] merchants. In the future, it is more suitable as a supplement
[00:11:14] to real shooting materials, rather than completely replacing real shooting.
[00:11:16] Let's also look at international
[00:11:20] news. After September, European Commission President von der Leyen first publicly
[00:11:24] mentioned promoting Canada to become the EU's first associate member.
[00:11:28] Local time on October 29th, Canada and the EU held a two-day
[00:11:31] EU-Canada summit in Montreal to further discuss deepening
[00:11:35] bilateral relations. However, just before the meeting, a
[00:11:39] video claiming to record a riot breaking out in Montreal was widely circulated on overseas social media.
[00:11:43] The post claimed that this was the Canadian version of European protest activities that occurred in Montreal.
[00:11:47] However, AFP fact-checked and found that the video was
[00:11:51] edited.
[00:12:11] Others fled in the chaos.
[00:12:12] 。
[00:12:16] Multiple posts publishing this video incorrectly identified the location as Montreal. However, AFP
[00:12:19] Fact Check reported that this video was actually
[00:12:23] filmed in 2024.
[00:13:10] However, it is worth noting that Montreal currently has experienced some small-scale protests,
[00:13:14] but the overall situation remains relatively peaceful. AFP
[00:13:17] also analyzed in its report that as Canada seeks closer
[00:13:21] ties with the EU, some Canadian critics.
[00:13:48] The CMG World Robot Mount Tai Competition was held in Tai'an,
[00:13:51] Shandong. The competition relies on the complex mountain terrain of Mount Tai.
[00:14:25] 640 meters, 31 steps.
[00:14:28] This section of the track has steps of varying heights and widths, and the
[00:14:32] slope changes frequently. It also has complex interferences such as obstacles and moss.
[00:14:36] I am currently on a section of the track, and my visual estimation
[00:14:39] is that the slope here can reach about 50 degrees, which is quite steep.
[00:14:43] And these steps are almost all different.
[00:14:47] Not only are the height and width different, but the flatness of the surface also
[00:14:51] varies greatly. So its height and width can be.
[00:17:07] At one point, according to the height of the steps and the slipperiness of the road surface, it constantly adjusted
[00:17:11] its foot placement and center of gravity. Robots need algorithms and hardware
[00:17:15] to coordinate and cooperate. Foot sensors detect impact forces when landing
[00:17:19] in real-time. Leg joints rely on compliant control
[00:17:23] to buffer forces. At the same time, algorithms pre-adjust
[00:17:26] the trunk posture to plan the landing position. The third test
[00:17:30] is energy consumption. Continuous climbing, high-frequency sensing, and repeated leveling.
[00:17:34] Every step mobilizes motors and computing power, which is very
[00:17:47] At the same time, it optimizes the planning of steps, reduces unnecessary posture oscillations and redundant movements, and saves
[00:17:51] precious battery power.
[00:17:55] On November 1st, the Chinese Academy of Sciences Institute of Vertebrate Paleontology and Paleoanthropology
[00:17:58] transferred a replica of the head and neck of the new Chuanmamenchisaurus
[00:18:02] to the Shanghai Natural History Museum, accompanying Shanghai citizens
[00:18:06] for 54 years.
[00:18:15] This is the new skull model of the Chuanmamenchisaurus. The skull is narrow and
[00:18:18] tall, and the teeth are spoon-shaped. It was replicated based on the characteristics of the
[00:18:22] Chuanmamenchisaurus fossil skull. However, the skull model of the
[00:18:26] Chuanmamenchisaurus previously exhibited at the Shanghai Natural History Museum was low and
[00:18:30] flat, and the teeth were rod-shaped, which was quite different from the
[00:18:33] Chuanmamenchisaurus skulls discovered in current archaeological findings.
[00:18:38] So why is there such a large difference? It is understood
[00:18:41] that the replica of the Chuanmamenchisaurus exhibited at the Shanghai Natural History Museum for 54
[00:18:45] years was replicated one-to-one in the early 1970s
[00:18:49] based on the fossil skeleton discovered in Hechuan, Chongqing, in 1957.
[00:18:53] Due to the missing skull of the original fossil skeleton at that time,
[00:18:57] biologists at the time referred to the morphology of the North American Diplodocus
[00:19:01] and fitted the Chuanmamenchisaurus skeleton on display with a replica
[00:19:04] of the Diplodocus skull. With the continuous discovery
[00:19:08] of Mamenchisaurus fossils with skulls in Sichuan, Xinjiang, and other places, the academic
[00:19:12] community gradually clarified the characteristics of the Mamenchisaurus skull.
[00:19:16] Therefore, the Shanghai Natural History Museum invited the Chinese Academy of Sciences
[00:19:20] Institute of Vertebrate Paleontology and Paleoanthropology to carry out scientific restoration,
[00:19:23] and finally restored a new skull model that matches
[00:19:27] the Chuanmamenchisaurus.
[00:19:31] 头 骨 模 型 。
[00:19:35] Based on surface scanning and CT scanning.
[00:20:07] Progress. On the other hand, we also hope to present the latest scientific and technological achievements
[00:20:11] in a very good way to the public.
[00:20:14] Recently, in the Wengjiatun
[00:20:18] area of the Caohai National Nature Reserve in Guizhou, staff monitored
[00:20:22] five black-necked cranes. This is the first batch of black-necked cranes to arrive at Caohai this year.
[00:20:26] Black-necked cranes are a national first-class protected wild animal. This batch of
[00:20:30] black-necked cranes belongs to the eastern population. Every year in late October, they depart from
[00:20:34] their breeding grounds on the plateau, such as Ruoergai in Sichuan and Gannan Maqu.
[00:21:54] According to statistics from Beijing Customs, in the first three quarters of this year, the total import and export value
[00:21:58] of service trade in the Beijing area was 2.78 trillion yuan, a year-on-year
[00:22:01] increase of
[00:22:02] 15.7%.
[00:22:08] It has achieved year-on-year growth for seven consecutive months, and the scale has reached a new high for the same period in history. Beijing's trade with 133 countries and regions around the world achieved
[00:22:12] positive growth. Local time
[00:22:15] on November 1st, the Chinese Navy's "Peace Ark" hospital ship, which is performing
[00:22:19] Mission Harmony 2026.
[00:22:26] This is the first time a Chinese naval hospital ship has visited Togo. During the visit, the task force will provide medical services
[00:22:30] on the main platform of the hospital ship and send out an advance team
[00:22:33] to visit and exchange views with the Togolese Naval Forces Hospital.
[00:22:37] The Spanish government announced on November 2nd that
[00:22:41] a high-speed train collision occurred in Toledo province in central Spain on October 28th.
[00:22:45] The identities of 66 of the 69 victims have been confirmed. The announcement showed
[00:22:49] that among the 66 confirmed victims, 54
[00:22:53] were Spanish nationals, four were American nationals,
[00:22:56] and one each from Argentina, Canada, Czech Republic,
[00:23:00] France, Japan, Mexico, Portugal, and Tunisia.
[00:23:04] A small
[00:23:07] plane crashed into a residential area in Polk County, Florida, on November 1st.
[00:23:10] All nine people on board died. According to
[00:23:14] local media reports, the plane descended more than 8,000 feet in just
[00:23:18] one minute. A large fire broke out near a main
[00:23:22] road, and about 600 people on the ground needed to be evacuated.
[00:23:25] Minutes before the crash, the pilot reported that the plane was icing
[00:23:29] and encountered moderate turbulence. Media reported that there were
[00:23:32] scattered thunderstorms in the affected area at that time. According to US media
[00:23:36] reports, on November 1st, the southwestern United States continued to experience heavy rainfall,
[00:23:40] causing flooding and prompting mandatory evacuations in many areas. In parts of Arizona
[00:23:44] and Colorado, flooding at one point caused people to be trapped
[00:23:47] and traffic to be disrupted. In addition,
[00:24:20] Beijing: Sunny, 4 to 19 degrees. Tokyo: Sunny to cloudy, 12 to 19
[00:24:24] degrees.
[00:24:26] Bangkok: Sunny to light rain,
[00:24:30] 27 to 36 degrees.
[00:24:33] Singapore: Cloudy to light rain, 26 to 33 degrees.
[00:24:36] Karachi: Sunny, 22 to 35
[00:24:40] degrees. Moscow: Cloudy to light
[00:24:43] rain, 0 to 5 degrees.
[00:24:47] Cairo: Sunny, 18 to 28 degrees.
[00:24:52] Paris: Light rain to clear, 7 to 11 degrees.
[00:24:54] 度
[00:24:56] London: Cloudy to light rain, 9 to 13 degrees.
[00:24:59] Sydney: Light rain, 19 to
[00:25:03] 26 degrees.
[00:25:05] Ottawa: Light rain, 9 to 12 degrees.
[00:25:08] New York: Sunny to cloudy, 11
[00:30:35] Hello, everyone. This is CCTV News Channel. At 1:00 AM, welcome to this
[00:30:38] news live broadcast. The reporter learned from the National Medical
[00:30:42] Insurance Administration on November 2nd that in the first nine months of this year, maternity
[00:30:46] insurance fund expenditures amounted to 11 billion yuan, covering
[00:30:50] 33.35 million people. Among them, maternity allowance
[00:30:53] expenditures amounted to 78.161 billion yuan, fully achieving direct payment
[00:30:57] to individuals.