In-vehicle atmosphere light effect generation method, storage medium and electronic device

Abstract

This invention discloses a method, storage medium, and electronic device for generating in-vehicle ambient lighting effects, comprising: acquiring historical interaction information of a target user and determining the user type of the target user; classifying multiple lighting effects into favorite lighting effects and cold start lighting effects based on the number of operations and a preset threshold; employing a content-based recommendation algorithm and a collaborative filtering recommendation algorithm to calculate the content-based score and collaborative filtering score of each cold start lighting effect, respectively; obtaining a first weight and a second weight according to the user type; for each cold start lighting effect, using the sum of the product of the content-based score and the first weight and the product of the collaborative filtering score and the second weight as a comprehensive score, and sorting the cold start lighting effects from high to low according to the comprehensive score; and obtaining the recommendation ratio corresponding to the user type to determine a recommended lighting effect list. This invention supports users in generating multiple types of lighting effects, avoiding the occurrence of information cocoons.

Description

Technical Field

[0001] This invention relates to the field of automotive technology, and in particular to a method for generating ambient lighting effects inside a vehicle, a storage medium, and an electronic device. Background Technology

[0002] With the continued rapid development of the automotive industry, the demand for emotional experiences in smart cockpits has significantly increased, highlighting the growing importance of ambient lighting. Ambient lighting not only provides basic illumination but also enhances the user experience by offering personalized emotional value through lighting effects and color combinations.

[0003] However, while the richness of lighting effects has significantly improved, preset lighting effects are insufficient to meet users' personalized needs. Current lighting effect customization faces multiple challenges: the operation is complex and lacks intuitiveness, and the professional barriers to color matching make it difficult to generate aesthetically pleasing lighting designs.

[0004] In existing technologies, traditional ambient light color selectors provide a color wheel for users to choose colors, but this has the following drawbacks: 1) Users cannot easily select the desired hue and saturation accurately, requiring high operational precision; 2) Although different colors can be selected for different areas to match the overall color scheme of the cabin, there are no color matching schemes provided or referenced. Users often just match randomly without a concrete concept of color matching, resulting in poor color matching effect. 3) Existing partitioning modes are either monotonous or cumbersome to operate. When combined with color matching, relying solely on custom generation results in colors lacking concrete meaning, which greatly reduces user experience and interest. 4) The existing automatic recommendation or random switching lighting effects function has limited effect and does not fully utilize the diverse functions of ambient lighting.

[0005] In terms of recommendation systems, existing in-vehicle ambient lighting systems typically lack effective user profiling mechanisms, resulting in simplistic recommendations that can create information cocoons. This means the system repeatedly recommends lighting effects similar to a user's historical preferences, preventing the user from encountering new lighting types. Furthermore, the sheer number of theoretical color combinations—for example, approximately 16.8 million single colors in the RGB color space, and a cube-sized number of three-color combinations—makes the feature vectors too sparse for direct similarity analysis and recommendation calculations. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method for generating in-vehicle ambient lighting effects, a storage medium, and an electronic device. This invention enables a brand-new interactive interface for emotional ambient lighting, supports users in generating various types of lighting effects, and combines a hybrid recommendation method based on content and collaborative filtering to effectively solve pain points such as the difficulty in generating lighting effects and the cumbersome nature of custom operations.

[0007] The technical solution of the present invention provides a method for generating in-vehicle ambient lighting effects, including: Obtain the target user's historical interaction information, which includes the number of times multiple lighting effects were operated; The user type of the target user is determined based on the historical interaction information; Based on the number of operations and the preset threshold number, the multiple lighting effects are divided into favorite lighting effects and cold start lighting effects; Content-based recommendation algorithm and collaborative filtering recommendation algorithm are used to calculate the content-based score and collaborative filtering score of each cold start lighting effect. Based on the user type, obtain the first weight of the content-based recommendation algorithm and the second weight of the collaborative filtering recommendation algorithm, wherein different user types correspond to different combinations of the first weight and the second weight; For each cold start lighting effect, the sum of the product of the content score and the first weight and the product of the collaborative filtering score and the second weight is used as a comprehensive score, and the cold start lighting effects are sorted from high to low according to the comprehensive score; Based on the user type, obtain the recommended matching ratio corresponding to the user type; Based on the recommended ratio, a list of recommended lighting effects is determined from the preferred lighting effects and the sorted cold start lighting effects.

[0008] Further, determining the user type of the target user based on the historical interaction information includes: Obtain the target user's operational behavior metrics and color usage metrics; The sum of the operational behavior index and the color usage index is calculated to obtain the user generalization score; The user type is determined based on the user generalization score and the preset grading threshold.

[0009] Furthermore, the operation behavior index is calculated based on the number of operation types used by the target user and the corresponding preset first score. The operation types include at least one of image generation lighting effects, emotion color selector color selection, theme lighting effect usage, partition mode setting, and lighting effect saving.

[0010] Furthermore, the color usage index is obtained using the following method: The continuous hue space is discretized into a first preset number of primary hues, and the continuous saturation space is discretized into a second preset number of saturation levels. Construct color regions multiplied by the first preset number and the second preset number; Statistically analyze the distribution of lighting effect colors historically used by the target user within the specified color area; The product of the number of hit color areas and the preset second score is used as the color usage index.

[0011] Furthermore, the step of classifying the multiple lighting effects into preferred lighting effects and cold start lighting effects based on the number of operations and a preset threshold number includes: If the number of operations is greater than or equal to the threshold number, the lighting effect is classified as the preferred lighting effect; If the number of operations is less than the threshold number, the lighting effect is classified as the cold start lighting effect.

[0012] Furthermore, the content-based recommendation algorithm and collaborative filtering recommendation algorithm are used to calculate the content-based score and collaborative filtering score for each cold start lighting effect, respectively, including: Extract the feature dimensions of the lighting effect, which include image tags, number of colors, partitioning method, and dynamic effects; Calculate the dimensional similarity between the cold start lighting effect and the preferred lighting effect under each of the aforementioned feature dimensions; The overall similarity is obtained by weighted summation of the similarities in each dimension. Using the number of operations corresponding to each of the preferred lighting effects as weights, the weighted average of the overall similarity of each cold start lighting effect is calculated to obtain the content-based score.

[0013] Furthermore, the content-based recommendation algorithm and collaborative filtering recommendation algorithm are used to calculate the content-based score and collaborative filtering score for each cold start lighting effect, respectively, including: Construct a user-lighting effect operation matrix, where the matrix elements represent the number of times each user operates on each lighting effect; Calculate the user similarity between the target user and other users; Select the user with the highest similarity to the target user as the neighbor user; Based on the number of times neighboring users interacted with the cold start lighting effect and the user similarity, the target user's interest in the cold start lighting effect was calculated, and the collaborative filtering score was obtained.

[0014] Further, calculating the user similarity between the target user and other users includes: The user similarity is calculated using the following formula: , in, For the target user With another user Among the various lighting effects that have been used together, take the smaller value of the number of operations for each and sum them up; This is the sum of the number of times the target user interacts with all lighting effects; This is the sum of the number of times another user performed operations on all lighting effects.

[0015] Further, the step of calculating the target user's interest in the cold start lighting effect based on the number of operations performed by neighboring users on the cold start lighting effect and the user similarity, to obtain the collaborative filtering score, includes: The collaborative filtering score is calculated using the following formula: , in, The collaborative filtering score; For the target user; For the aforementioned cold start lighting effect; In order to connect with the target user The neighboring users with the highest similarity; For those who have used the aforementioned cold start lighting effect Users; For the target user With another user User similarity; For another user For the cold start lighting effect Number of operations.

[0016] Further, the step of determining a recommended lighting effect list from the preferred lighting effects and the sorted cold start lighting effects based on the recommended ratio includes: Based on the recommended ratio, determine the recommended quantity of the preferred lighting effects and the recommended quantity of the cold start lighting effects; Select a corresponding number of lighting effects from the preferred lighting effects, and then select the top-ranked lighting effects from the sorted cold start lighting effects to obtain the recommended lighting effects list.

[0017] Furthermore, after obtaining the recommended lighting effect list, the process further includes: If the recommended lighting effect is the preferred lighting effect, a color scheme is generated based on the target user's historical color preferences; If the recommended lighting effect is the cold start lighting effect, obtain the color preferences of similar users who have operated on the lighting effect the most, and generate a color scheme based on the color preferences of similar users.

[0018] Furthermore, the color preference generation color scheme includes: Prioritizing saturation: The color space is divided into a first preset number of large regions according to saturation levels. The region with the highest hit frequency and that meets the image requirements of the lighting effect is selected as the target region. After further subdividing the target region according to hue, a clustering algorithm is used for cluster analysis. Hue priority: The color space is divided into a second preset number of large regions according to the main color types. The region with the highest hit frequency and that meets the image requirements of the lighting effect is selected as the target region. After splitting the target region according to the subdivided hue saturation, a clustering algorithm is used for cluster analysis.

[0019] The present invention also provides a computer-readable storage medium that stores computer instructions, which, when executed by a computer, are used to perform all the steps of the in-vehicle ambient lighting effect generation method described above.

[0020] The present invention also provides an electronic device, comprising: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which enables the at least one processor to perform the in-vehicle ambient lighting generation method as described above.

[0021] The above technical solution has the following beneficial effects: By acquiring historical interaction information of target users, the user type is determined based on this information. Multiple lighting effects are then categorized into preferred effects and cold-start effects based on the number of operations. A hybrid recommendation algorithm dynamically weights and combines content-based recommendations with collaborative filtering recommendations, adaptively adjusting the weight ratio of the two algorithms based on user type. This achieves a dynamic balance between accurately recommending user-preferred lighting effects and exploring new lighting effects, ensuring the relevance of recommendations while expanding users' aesthetic horizons and providing more diverse lighting effects, thus avoiding information cocoons. Furthermore, through the recommended lighting effect list, users can automatically obtain aesthetically pleasing lighting effect color schemes simply by selecting concrete color image tags in the graphical user interface, without needing professional color matching knowledge, thus improving versatility. Simultaneously, the continuous hue space is discretized into a finite number of primary hues, and the continuous saturation space is discretized into a finite number of levels, constructing a finite number of discrete color regions. This reduces the dimensionality of the color space from tens of millions to dozens of regions, significantly reducing the sparsity of feature vectors and the complexity of subsequent similarity calculations. Furthermore, clustering algorithms are used in the color scheme generation process to automatically remove discrete color points and merge similar color clusters, avoiding the generation of abrupt and uncoordinated color schemes and ensuring the aesthetic consistency of the output color schemes. Attached Figure Description

[0022] The disclosure of this invention will become more readily understood by referring to the accompanying drawings. It should be understood that these drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. In the drawings: Figure 1 A flowchart illustrating a method for generating in-vehicle ambient lighting effects according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the GUI color image selector interface in an embodiment of the present invention; Figure 3 This is a schematic diagram of a color image selector according to an embodiment of the present invention; Figure 4 This is a flowchart illustrating the user type determination algorithm in an embodiment of the present invention. Figure 5 This is a schematic diagram of color space discretization in an embodiment of the present invention; Figure 6 This is a flowchart illustrating the workflow of the content-based recommendation algorithm in this embodiment of the invention. Figure 7 This is a flowchart illustrating the workflow of the collaborative filtering recommendation algorithm in this embodiment of the invention. Figure 8 A flowchart illustrating a method for generating ambient lighting effects in a vehicle, provided as a preferred embodiment of the present invention; Figure 9 This is a schematic diagram of the hardware structure of an electronic device for generating ambient lighting effects inside a vehicle, provided as an embodiment of the present invention. Detailed Implementation

[0023] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0024] It is readily understood that, based on the technical solution of this invention, various structural and implementation methods can be interchanged by those skilled in the art without altering the essential spirit of the invention. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this invention and should not be considered as the entirety of the invention or as limitations or restrictions on the technical solution of the invention.

[0025] The directional terms such as up, down, left, right, front, back, front, back, top, and bottom mentioned or possibly used in this specification are defined relative to the structures shown in the accompanying drawings. They are relative concepts and may therefore vary depending on their location and usage. Therefore, these or other directional terms should not be interpreted as restrictive.

[0026] like Figure 1 As shown, an embodiment of the present invention provides a method for generating ambient lighting effects in a vehicle interior, comprising: Step S101: Obtain the target user's historical interaction information, which includes the number of times multiple lighting effects were operated; Step S102: Determine the user type of the target user based on the historical interaction information; Step S103: Based on the number of operations and the preset threshold number, divide the multiple lighting effects into favorite lighting effects and cold start lighting effects; Step S104: Using content-based recommendation algorithm and collaborative filtering recommendation algorithm, calculate the content-based score and collaborative filtering score of each cold start lighting effect respectively; Step S105: Based on the user type, obtain the first weight of the content-based recommendation algorithm and the second weight of the collaborative filtering recommendation algorithm, wherein different user types correspond to different combinations of the first weight and the second weight; Step S106: For each cold start lighting effect, the sum of the product of the content score and the first weight and the product of the collaborative filtering score and the second weight is used as the comprehensive score, and the cold start lighting effects are sorted from high to low according to the comprehensive score; Step S107: Based on the user type, obtain the recommended matching ratio corresponding to the user type; Step S108: Based on the recommended ratio, determine a list of recommended lighting effects from the preferred lighting effects and the sorted cold start lighting effects.

[0027] Specifically, this invention can be applied to electronic devices with processing capabilities, such as vehicle controllers. For example, the Electronic Control Unit (ECU) of a vehicle.

[0028] First, execute step S101 to obtain the target user's historical interaction information.

[0029] Specifically, historical interaction information refers to historical data generated after a target user operates the in-vehicle ambient lighting system through a graphical user interface (GUI) for a period of time. Historical interaction information includes the number of times multiple lighting effects were operated. For example... Figure 2 As shown, the GUI interface includes a partition operation area, a color operation area, an image operation area, an automatic recommendation module, and an image generation module. The GUI interface can be customized according to user needs. Figure 3 As shown, setting different color images in the GUI will display a variety of preset color images (such as "New Oriental", "Charm of Tea Ceremony", "Elegance", "Avant-garde", "Warmth", "Unity", "Rhythm" etc.). Each color image is matched with a corresponding color library, which contains the preset number of colors, hue range and saturation range. After the user selects the target color image, the system automatically generates lighting effect parameters based on the corresponding color library and applies them to the cabin ambient lighting.

[0030] Next, step S102 is executed to determine the user type of the target user based on historical interaction information.

[0031] Specifically, the user type of the target user can be determined based on historical interaction information and empirical values, or other existing technologies can be used to determine the user type. In this embodiment, the user type is divided into three categories (level 1 / 2 / 3). The higher the level, the more likely the user is to accept new lighting effects. Therefore, in the recommended lighting effects, more lighting effects that have not been used before (i.e., cold start lighting effects) can be recommended, and more frequently used lighting effects (i.e., favorite lighting effects) can be recommended to increase the sense of novelty.

[0032] In one embodiment, such as Figure 4 As shown, determining the user type of the target user based on the historical interaction information includes: Step S401: Obtain the target user's operation behavior indicators and color usage indicators; Step S402: Calculate the sum of the operation behavior index and the color usage index to obtain the user generalization score; Step S403: Determine the user type based on the user generalization score and the preset grading threshold.

[0033] Specifically, taking the virtual user "Xiaoming" as an example, the execution process of determining the user type is explained in detail: Retrieve Xiaoming's historical operation records and count the number of types and scores for each operation, as shown in Table 1 below: Table 1

[0034] The original RGB color space theoretically has 256 colors. 3 =16,777,216 possibilities, the number of combinations of three colors reaches 4.7 × 10⁻⁶. 21 The feature vectors are too sparse to be directly used for similarity calculation due to their large size. Therefore, this application discretizes the 360° continuous hue ring into 24 main hues (e.g., ...). Figure 5 As shown, each 15° represents a hue interval. The continuous saturation from 0-100% is discretized into three levels (low saturation, medium saturation, and high saturation). After dimensionality reduction, 72 discrete color regions are obtained (24 primary hues × 3 saturation levels). N×M color regions are constructed, and all historical lighting effect colors used by the user are mapped to their corresponding positions within these 72 regions. The distribution of historical lighting effect colors used in each hour across these regions is statistically analyzed, and 18 regions are matched. Therefore, the color usage index = 18 × 0.5 = 9 points.

[0035] User generalization score = Operational behavior index + Color usage index = 42.5 + 9 = 51.5 points.

[0036] The user's generalization score was compared with the grading threshold, as shown in Table 2 below: Table 2

[0037] Xiaoming's user generalization score is 51.5 points, which, according to Table 2, belongs to Level 2 (General).

[0038] In one embodiment, the operation behavior index is calculated based on the number of operation types used by the target user and the corresponding preset first score. The operation types include at least one of image generation lighting effects, emotion color selector color selection, theme lighting effect usage, partition mode setting, and lighting effect saving.

[0039] In one embodiment, the color usage index is obtained using the following method: The continuous hue space is discretized into a first preset number of primary hues, and the continuous saturation space is discretized into a second preset number of saturation levels. Construct color regions multiplied by the first preset number and the second preset number; Statistically analyze the distribution of lighting effect colors historically used by the target user within the specified color area; The product of the number of hit color areas and the preset second score is used as the color usage index.

[0040] Then, step S103 is executed, and multiple lighting effects are divided into favorite lighting effects and cold start lighting effects according to the number of operations and the preset number threshold.

[0041] Specifically, the number of operations corresponding to each light effect is compared with a preset threshold. If the number of operations is greater than or equal to the threshold, the light effect is classified as a "favorite light effect"; if the number of operations is less than the threshold, the light effect is classified as a "cold start light effect". Continuing with Xiaoming as an example, with a threshold of 30 operations, Xiaoming has used the following light effects (sorted by number of operations), as shown in Table 3 below: Table 3

[0042] As can be seen from Table 3 above, Xiaoming's favorite lighting effects (operation times greater than or equal to 30) are A1, A2, and A3, and the cold start lighting effects (operation times less than 30) are A4, A5, and A6.

[0043] "Cold start lighting effects" refer to lighting effects that users have never used or whose number of operations is less than a threshold (e.g., set to 30 times). Because these lighting effects lack user behavior data, it is impossible to directly determine whether users like them based on historical preferences. Therefore, it is necessary to use two methods, "content-based" and "collaborative filtering," to predict users' potential interests.

[0044] Next, step S104 is executed, using a content-based recommendation algorithm and a collaborative filtering recommendation algorithm to calculate the content-based score and collaborative filtering score of each cold start lighting effect. Specifically, the core idea of ​​content-based recommendation algorithms is that if lighting effect A is a favorite lighting effect and lighting effect B is a cold start lighting effect, and calculations show that lighting effect B has a high similarity to lighting effect A, then it is expected that the user will also like lighting effect B. Therefore, lighting effect B can be used as a highly recommended lighting effect in the cold start lighting effect category.

[0045] The core idea of ​​collaborative filtering recommendation algorithms is that if user A is very similar to user B, then it is expected that user B will also like the lighting effects that user A likes.

[0046] In one embodiment, such as Figure 6 As shown, the content-based recommendation algorithm and collaborative filtering recommendation algorithm are used to calculate the content-based score and collaborative filtering score of each cold start lighting effect, respectively, including: Step S601: Extract the feature dimensions of the lighting effect, including image tags, number of colors, partitioning method, and dynamic effects; Step S602: Calculate the dimensional similarity between the cold start lighting effect and the preferred lighting effect in each of the feature dimensions; Step S603: Perform a weighted summation of the similarities of each dimension to obtain the overall similarity; Step S604: Using the number of operations corresponding to each of the preferred lighting effects as weights, calculate the weighted average of the overall similarity of each cold start lighting effect to obtain the content-based score.

[0047] Specifically, taking the cold start lighting effect A6 (elegant, single color throughout the cabin) in Table 3 above as an example, the calculation process based on content score is explained: The image of A6 is "refined and elegant," with image information including low saturation, green hue (H140), and low contrast. The image of A1 is also "refined and elegant," with image information including low saturation, green hue (H120), and low contrast.

[0048] First, based on the images and image information of A6 and A1, a weighted sum is calculated for image tags, hue difference, saturation, and contrast to obtain the image similarity between A6 and A1. Then, based on the image similarity, the number of colors, partitioning method, and dynamic effects are added as dimensions, and the similarity of each dimension is weighted and summed to obtain the overall similarity between A6 and A1.

[0049] Similarly, calculate the overall similarity between A6 and A2 and A3: Sim(A6, A2) = 0.42 (Warm and cozy vs. elegant, different images, significant difference); Sim(A6, A3) = 0.51 (Unity vs. elegance, both are gentle but have different images).

[0050] The content-based score calculation for the cold start lighting effect A6: Score(A6) = Σ[ Sim(A6, Ai) × Number of uses (Ai) ] / Σ Number of uses (Ai) = (0.744×85 + 0.42×60 + 0.51×40) / (85+60+40) = (63.24 + 25.2 + 20.4) / 185 = 108.84 / 185 = 0.588 Similarly, the content-based scores for other cold start lighting effects are calculated as shown in Table 4 below: Table 4

[0051] As can be seen from Table 4 above, the content-based score of the cold start lighting effect A6 is 0.588, ranking first, followed by A4 and A5.

[0052] In this embodiment, a content-based recommendation algorithm is used to calculate the similarity between cold start lighting effects and user-favorite lighting effects in dimensions such as appearance, color scheme, and zoning, and to recommend lighting effects in various ways, thereby generating a variety of lighting effects and increasing the diversity of ambient lighting effects.

[0053] In one embodiment, such as Figure 7 As shown, the content-based recommendation algorithm and collaborative filtering recommendation algorithm are used to calculate the content-based score and collaborative filtering score of each cold start lighting effect, respectively, including: Step S701: Construct a user-lighting effect operation matrix, where the matrix elements represent the number of times each user operates on each lighting effect; Step S702: Calculate the user similarity between the target user and other users; Step S703: Select the user with the highest similarity to the target user as the neighbor user; Step S704: Based on the number of times the neighboring users operate on the cold start lighting effect and the user similarity, calculate the target user's interest in the cold start lighting effect and obtain the collaborative filtering score.

[0054] Specifically, assuming there are three other users B, C, and D in the system, construct a complete user-lighting effect operation matrix, where the matrix elements represent the number of times each user operates on each lighting effect, as shown in Table 5 below: Table 5

[0055] Taking Xiaoming and User B as an example, the lighting effects they used together are A1, A2, A3, A4, and A5. For each shared lighting effect, take the smaller value of the number of operations: A1: min(85,70)=70, A2: min(60,50)=50, A3: min(40,35)=35, A4: min(12,3)=3, A5: min(5,2)=2; The intersection |N(Xiaoming)∩N(B)| = 70+50+35+3+2 = 160; |N(Xiaoming)| = 85+60+40+12+5+0 = 202; |N(B)| = 70+50+35+3+2+45 = 205; W(Xiaoming, B) = = 160 / 203.5 = 0.786.

[0056] Similarly: W(Xiaoming,C) = 0.31, W(Xiaoming,D) = 0.72.

[0057] The preset user number threshold K=3 (i.e., the top 3 most similar users) is selected, and the user is ranked as follows: User B (0.786) > User D (0.72) > User C (0.31).

[0058] Taking the cold start lighting effect A6 as an example, N(A6) is the set of users who have performed operations on A6 = {user B, user C, user D} (the number of times they were used are 45, 5, and 50 respectively).

[0059] S(Xiaoming, K) ∩ N(A6) = {User B, User C, User D} (all 3 are present) p(Xiaoming, A6) = Σ W(Xiaoming, v) × Rv(A6) = 0.786×45 + 0.31×5 + 0.72×50 = 35.37 + 1.55 + 36.0 = 72.92 This allows us to determine Xiaoming's level of interest in the cold start lighting effect A6.

[0060] In this embodiment, a collaborative filtering recommendation algorithm is used to find similar user groups, recommend lighting effects that the group likes, and further generate various types of lighting effects to increase the diversity of ambient lighting effects.

[0061] In one embodiment, calculating the user similarity between the target user and other users includes: The user similarity is calculated using the following formula: , in, For the target user With another user Among the various lighting effects that have been used together, take the smaller value of the number of operations for each and sum them up; This is the sum of the number of times the target user interacts with all lighting effects; This is the sum of the number of times another user performed operations on all lighting effects.

[0062] In one embodiment, the step of calculating the target user's interest in the cold start lighting effect based on the number of operations performed by neighboring users on the cold start lighting effect and the user similarity, and obtaining the collaborative filtering score, includes: The collaborative filtering score is calculated using the following formula: , in, The collaborative filtering score; For the target user; For the aforementioned cold start lighting effect; In order to connect with the target user The neighboring users with the highest similarity; For those who have used the aforementioned cold start lighting effect Users; For the target user With another user User similarity; For another user For the cold start lighting effect Number of operations.

[0063] Then, step S105 is executed to obtain the first weight of the content-based recommendation algorithm and the second weight of the collaborative filtering recommendation algorithm based on the user type.

[0064] Specifically, the weights of the two routes are not fixed, but dynamically adjusted according to user type. Relaxed users tend to be conservative and trust the similarity of the lighting effects themselves (high weight based on content); tech users are more willing to try new things that others like (high weight based on collaborative filtering).

[0065] Then, step S106 is executed: for each cold start lighting effect, the sum of the product of the content score and the first weight and the product of the collaborative filtering score and the second weight is used as the comprehensive score, and the cold start lighting effects are sorted from high to low according to the comprehensive score.

[0066] Specifically, the content score and collaborative filtering score are normalized separately to obtain normalized G1 and G2 values. Xiaoming is user type 2 (general), corresponding to W1 = 0.5 and W2 = 0.5. Therefore, Xiaoming's overall score = G1 × W1 + G2 × W2. Thus, Xiaoming's overall scores for the cold start lighting effects A6, A5, and A4 are shown in Table 6 below: Table 6

[0067] The cold start lighting effects are then ranked from highest to lowest based on the overall score.

[0068] Then, step S107 is executed to obtain the recommended ratio corresponding to the user type.

[0069] Specifically, after calculating the overall score, the ranking results are checked to see if any lighting effects have been marked as negative feedback by the user. For example, if Xiaoming long-presses "Next" in automatic recommendation mode and passes by the "Pioneer Dual-Color Straight Side Display" lighting effect, this effect is added to the filter list and removed from the cold start candidate set. As shown in Table 6 above, Xiaoming dislikes "Pioneer Dual-Color Straight Side Display," while A4 shows "Pioneer Dual-Color Front and Back." Although both feature Pioneer images, they are in different zones and are not on the dislike list, so they are retained.

[0070] Then, the preset recommended ratios corresponding to user types are obtained. As shown in Table 2 above, Xiaoming belongs to the general type, and the corresponding recommended ratio is 6:4.

[0071] Finally, step S108 is executed to determine a recommended lighting effect list from the preferred lighting effects and the sorted cold start lighting effects according to the recommended ratio.

[0072] Specifically, based on the recommended ratio of 6:4 corresponding to Xiaoming's user type, 6 lighting effects are recommended from the favorite lighting effects, and 4 lighting effects are recommended from the sorted cold start lighting effects, for a total of 10.

[0073] Taking Xiaoming as an example, the selection of Xiaoming's favorite lighting effects is as follows: instead of directly taking all A1, A2, and A3, six lighting effects are randomly generated from Xiaoming's favorite lighting effect pool to ensure that each recommendation is fresh and not completely repetitive. The selection of Xiaoming's cold start lighting effects: as shown in Table 3, since there are only three cold start lighting effects in this embodiment, A6, A4, and A5 are all taken out as the recommended lighting effect list.

[0074] Xiaoming's final list of recommended lighting effects consists of 6 favorite lighting effects and 3 cold start lighting effects.

[0075] like Figure 8 As shown, a preferred embodiment of the present invention provides a method for generating in-vehicle ambient lighting effects, comprising: Step S801: Obtain the target user's historical interaction information; Step S802: Determine the user type of the target user based on historical interaction information; Step S803: Based on the number of operations and the preset threshold number of operations, divide the multiple lighting effects into favorite lighting effects and cold start lighting effects; Step S804: Using content-based recommendation algorithm and collaborative filtering recommendation algorithm, calculate the content-based score and collaborative filtering score of each cold start lighting effect respectively; Step S805: Based on the user type, obtain the first weight of the content-based recommendation algorithm and the second weight of the collaborative filtering recommendation algorithm; Step S806: For each cold start lighting effect, the sum of the product of the content score and the first weight and the product of the collaborative filtering score and the second weight is used as the comprehensive score, and the cold start lighting effects are sorted from high to low according to the comprehensive score. Step S807: Based on the user type, obtain the recommended matching ratio corresponding to the user type; Step S808: Based on the recommended ratio, determine the recommended number of preferred lighting effects and the recommended number of cold start lighting effects; Step S809: Select the corresponding number of lighting effects from the favorite lighting effects, and select the corresponding number of top-ranked lighting effects from the sorted cold start lighting effects to obtain the recommended lighting effect list; Step S810: If the recommended lighting effect is the preferred lighting effect, generate a color scheme based on the target user's historical color preferences; Step S811: If the recommended lighting effect is the cold start lighting effect, obtain the color preferences of similar users who have operated on this lighting effect the most, and generate a color scheme based on the color preferences of similar users; Specifically, for each lighting effect in the recommended lighting effect list, a specific color scheme needs to be generated. The source of color data depends on the type of lighting effect: preferred lighting effects directly use the target user's own historical color preference data; cold start lighting effects are obtained from the color preference data of the user with the most operations among similar users in collaborative filtering.

[0076] Taking Xiaoming as an example, for the preferred lighting effects: read Xiaoming's historical color preference data from lighting effects A1, A2, and A3; for the cold start lighting effects: find the lighting effect with the most user B operations (45 times) from lighting effect A6, and then read user B's color preference data.

[0077] In this embodiment, after obtaining the list of recommended lighting effects, a color scheme is generated for each lighting effect in the list to ensure that each recommendation is fresh and not completely repetitive.

[0078] In one embodiment, the color preference generation of a color scheme includes: Prioritizing saturation: The color space is divided into a first preset number of large regions according to saturation levels. The region with the highest hit frequency and that meets the image requirements of the lighting effect is selected as the target region. After further subdividing the target region according to hue, a clustering algorithm is used for cluster analysis. Hue priority: The color space is divided into a second preset number of large regions according to the main color types. The region with the highest hit frequency and that meets the image requirements of the lighting effect is selected as the target region. After splitting the target region according to the subdivided hue saturation, a clustering algorithm is used for cluster analysis.

[0079] Specifically, taking the cold start lighting effect A6 (elegant, single color throughout the cabin, N=1 color) as an example, the color matching is done using user B's color data.

[0080] Color matching follows two parallel approaches, emphasizing saturation and hue respectively: 1. Saturation-focused approach: The color space is divided into 3×3×3 = 27 regions based on high / medium / low saturation. The number of times user B uses the color space in each region is counted. The region Q with the highest frequency is found. Region Q is further subdivided into 24 hues. A clustering algorithm (Density-Based Spatial Clustering of Applications with Noise, DBSCAN) is used to remove discrete points and merge similar color clusters. The frequency of each point is combined to obtain the sub-region with the highest frequency. RGB values ​​are randomly generated within this region.

[0081] 2. Emphasis on hue: The color space is divided into 6×6×6 = 216 regions based on 6 primary colors. The number of times user B uses the color in each region is counted. The region J with the most usage is found. Region J is further subdivided by hue and saturation. DBSCAN clustering is used to obtain the sub-region with the most usage. RGB values ​​are then randomly generated within this region.

[0082] In this embodiment, color schemes are generated using both saturation and hue approaches, covering two different aesthetic preferences. Users sensitive to saturation can obtain color schemes through the saturation approach, while users sensitive to hue can obtain color schemes through the hue approach. Simultaneously, a clustering algorithm automatically removes outliers, preventing the generation of jarring and discordant color schemes while preserving the freshness of the color schemes.

[0083] One embodiment of the present invention provides a computer-readable storage medium for storing computer instructions, which, when executed by a computer, are used to perform all steps of the in-vehicle ambient lighting effect generation method as described in any of the above method embodiments.

[0084] like Figure 9 As shown, a hardware structure diagram of an electronic device for generating in-vehicle ambient lighting effects according to an embodiment of the present invention includes: At least one processor 901; and, A memory 902 is communicatively connected to at least one processor 901; wherein, The memory 902 stores instructions that can be executed by at least one processor 901, which enables the at least one processor 901 to perform the in-vehicle ambient lighting effect generation method as described in any of the above method embodiments.

[0085] Figure 9 Take the 901 processor as an example.

[0086] The electronic device is preferably an electronic control unit (ECU).

[0087] The electronic device may also include an input device 903 and an output device 904.

[0088] The processor 901, memory 902, input device 903 and output device 904 can be connected by a bus or other means. The figure shows an example of connection by bus.

[0089] The memory 902, as a non-volatile computer-readable storage medium, can be used to obtain non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the in-vehicle ambient lighting effect generation method in the embodiments of this application, for example, Figure 1 , Figure 4 , Figures 6-8 The method flow is shown. The processor 901 executes various functional applications and data processing by running non-volatile software programs, instructions, and modules acquired in the memory 902, thereby realizing the in-vehicle ambient lighting effect generation method in the above embodiment.

[0090] The memory 902 may include a program acquisition area and a data acquisition area, wherein the program acquisition area may acquire an operating system and an application program required for at least one function; the data acquisition area may acquire data created according to the use of the in-vehicle ambient lighting effect generation method, etc. Furthermore, the memory 902 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 902 may optionally include memory remotely located relative to the processor 901, and these remote memories may be connected via a network to the apparatus performing the in-vehicle ambient lighting effect generation method. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0091] The input device 903 can receive user clicks and generate signal inputs related to user settings and function control of the in-vehicle ambient lighting effect generation method. The output device 904 may include a display screen or other display device.

[0092] When the one or more modules are accessed in the memory 902 and are run by the one or more processors 901, the in-vehicle ambient lighting effect generation method in any of the above method embodiments is executed.

[0093] The above-described product can perform the methods provided in the embodiments of this application, and has the corresponding functional modules and beneficial effects for performing the methods. Technical details not described in detail in this embodiment can be found in the methods provided in the embodiments of this application.

[0094] The above embodiments are only used to illustrate the technical solutions of the embodiments of the present invention, and are not intended to limit them. Although the embodiments of the present invention have been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for generating ambient lighting effects inside a vehicle, characterized in that, include: Obtain the target user's historical interaction information, which includes the number of times multiple lighting effects were operated; The user type of the target user is determined based on the historical interaction information; Based on the number of operations and the preset threshold number, the multiple lighting effects are divided into favorite lighting effects and cold start lighting effects; Content-based recommendation algorithm and collaborative filtering recommendation algorithm are used to calculate the content-based score and collaborative filtering score of each cold start lighting effect. Based on the user type, obtain the first weight of the content-based recommendation algorithm and the second weight of the collaborative filtering recommendation algorithm, wherein different user types correspond to different combinations of the first weight and the second weight; For each cold start lighting effect, the sum of the product of the content score and the first weight and the product of the collaborative filtering score and the second weight is used as a comprehensive score, and the cold start lighting effects are sorted from high to low according to the comprehensive score; Based on the user type, obtain the recommended matching ratio corresponding to the user type; Based on the recommended ratio, a list of recommended lighting effects is determined from the preferred lighting effects and the sorted cold start lighting effects.

2. The method for generating in-vehicle ambient lighting effects as described in claim 1, characterized in that, Determining the user type of the target user based on the historical interaction information includes: Obtain the target user's operational behavior metrics and color usage metrics; The sum of the operational behavior index and the color usage index is calculated to obtain the user generalization score; The user type is determined based on the user generalization score and the preset classification threshold; Preferably, the operation behavior index is calculated based on the number of operation types used by the target user and the corresponding preset first score. The operation types include at least one of image generation lighting effects, emotion color selector color selection, theme lighting effect usage, partition mode setting, and lighting effect saving. Preferably, the color usage index is obtained using the following method: The continuous hue space is discretized into a first preset number of primary hues, and the continuous saturation space is discretized into a second preset number of saturation levels. Construct color regions multiplied by the first preset number and the second preset number; Statistically analyze the distribution of lighting effect colors historically used by the target user within the specified color area; The product of the number of hit color areas and the preset second score is used as the color usage index.

3. The method for generating in-vehicle ambient lighting effects as described in claim 1, characterized in that, The step of classifying multiple lighting effects into preferred lighting effects and cold start lighting effects based on the number of operations and a preset threshold number includes: If the number of operations is greater than or equal to the threshold number, the lighting effect is classified as the preferred lighting effect; If the number of operations is less than the threshold number, the lighting effect is classified as the cold start lighting effect.

4. The method for generating in-vehicle ambient lighting effects as described in claim 1, characterized in that, The method employs content-based recommendation algorithms and collaborative filtering recommendation algorithms to calculate the content-based score and collaborative filtering score for each cold start lighting effect, including: Extract the feature dimensions of the lighting effect, which include image tags, number of colors, partitioning method, and dynamic effects; Calculate the dimensional similarity between the cold start lighting effect and the preferred lighting effect under each of the aforementioned feature dimensions; The overall similarity is obtained by weighted summation of the similarities in each dimension. Using the number of operations corresponding to each of the preferred lighting effects as weights, the weighted average of the overall similarity of each cold start lighting effect is calculated to obtain the content-based score.

5. The method for generating in-vehicle ambient lighting effects as described in claim 1, characterized in that, The method employs content-based recommendation algorithms and collaborative filtering recommendation algorithms to calculate the content-based score and collaborative filtering score for each cold start lighting effect, including: Construct a user-lighting effect operation matrix, where the matrix elements represent the number of times each user operates on each lighting effect; Calculate the user similarity between the target user and other users; Select the user with the highest similarity to the target user as the neighbor user; Based on the number of times the neighboring users operated on the cold start lighting effect and the user similarity, the target user's interest in the cold start lighting effect is calculated, and the collaborative filtering score is obtained. Preferably, calculating the user similarity between the target user and other users includes: The user similarity is calculated using the following formula: , in, For the target user With another user Among the various lighting effects that have been used together, take the smaller value of the number of operations for each and sum them up; This is the sum of the number of times the target user interacts with all lighting effects; This is the sum of the number of times another user interacted with all the lighting effects; Preferably, the step of calculating the target user's interest in the cold start lighting effect based on the number of operations performed by neighboring users on the cold start lighting effect and the user similarity, to obtain the collaborative filtering score, includes: The collaborative filtering score is calculated using the following formula: , in, The collaborative filtering score; For the target user; For the aforementioned cold start lighting effect; In order to connect with the target user The neighboring users with the highest similarity; For those who have used the aforementioned cold start lighting effect Users; For the target user With another user User similarity; For another user For the cold start lighting effect Number of operations.

6. The method for generating in-vehicle ambient lighting effects as described in claim 1, characterized in that, The step of determining a recommended lighting effect list from the preferred lighting effects and the sorted cold start lighting effects according to the recommended ratio includes: Based on the recommended ratio, determine the recommended quantity of the preferred lighting effects and the recommended quantity of the cold start lighting effects; Select a corresponding number of lighting effects from the preferred lighting effects, and then select the top-ranked lighting effects from the sorted cold start lighting effects to obtain the recommended lighting effects list.

7. The method for generating in-vehicle ambient lighting effects as described in claim 6, characterized in that, After obtaining the recommended lighting effects list, the process further includes: If the recommended lighting effect is the preferred lighting effect, a color scheme is generated based on the target user's historical color preferences; If the recommended lighting effect is the cold start lighting effect, obtain the color preferences of similar users who have operated on the lighting effect the most, and generate a color scheme based on the color preferences of similar users.

8. The method for generating in-vehicle ambient lighting effects as described in claim 7, characterized in that, The color preference generation color scheme includes: Prioritizing saturation: The color space is divided into a first preset number of large regions according to saturation levels. The region with the highest hit frequency and that meets the image requirements of the lighting effect is selected as the target region. After further subdividing the target region according to hue, a clustering algorithm is used for cluster analysis. Prioritize hue: Divide the color space into a second preset number of large regions according to the main color categories, select the region with the highest hit frequency and that meets the image requirements of the lighting effect as the target region, and then perform cluster analysis using a clustering algorithm after splitting the target region according to the subdivided hue saturation.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, which, when executed by a computer, are used to perform all the steps of the in-vehicle ambient lighting effect generation method as described in any one of claims 1-8.

10. An electronic device, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the in-vehicle ambient lighting generation method as described in any one of claims 1-8.

Images